Tuesday, October 27, 2026 · Edition updated 6:00 AM ETTUE, OCT 27 editionNext edition in Next in --:--:--
Intelligence Verdict
Can AI do it yet? · School

Can AI write my essay or homework yet?

AI can already write exam answers that markers miss and that out-score real students, but a small MIT trial found people who wrote essays with ChatGPT engaged their brains less and did worse over time.

  • 4 of 6Tested with real people
  • 2pieces of evidence
The answer

Stage 4: tested with real people.

The rule that placed itA randomized trial or field study has tested it with real people doing the task.

The evidenceIn an MIT Media Lab study, people randomly assigned to write essays with ChatGPT showed the weakest brain connectivity and underperformed over four months. It is a preprint, not yet peer-reviewed. Kosmyna et al., MIT Media Lab, arXiv 2506.08872, Jun 10, 2025.

A stage says how far the evidence has got. It is not a rating of the AI and not advice.

  1. 1
    ClaimedNo evidence yet
  2. 2
    LaunchedNo evidence yet
  3. 3
    Independently testedEvidence found
  4. 4
    Tested with real peoplePlaced here
  5. 5
    Regulator-clearedNo approval process exists
  6. 6
    Widely used with a verified recordNo evidence yet
The evidence

Everything we found.

In ladder order, newest first within each kind. Makers' own claims are labeled as theirs.

Tested

What independent tests found

  • University of Reading (PLOS ONE)

    In a blind test of real university psychology exams, ChatGPT-written answers went unnoticed by markers 94% of the time and on average got higher grades than real students.

    “went undetected in 94% of cases and, on average, attained higher grades than real student submissions.”

    University of Reading · Jun 26, 2024

Real people

Tested with real people

  • Randomized trial, 54 participants (18 in a fourth session)

    In an MIT Media Lab study, people randomly assigned to write essays with ChatGPT showed the weakest brain connectivity and underperformed over four months. It is a preprint, not yet peer-reviewed.

    “Over four months, LLM users consistently underperformed at neural, linguistic, and behavioral levels.”

    Kosmyna et al., MIT Media Lab, arXiv 2506.08872 · Jun 10, 2025

Not found in our sources yet: what the makers claim; what has launched; how many people use it. No regulator approves AI for this task, so the approval step doesn't apply.

Beside the stage

Track record and tools.

Failures and court cases are shown beside the stage, never folded into it. Tools are listed because the evidence names them, in no order of merit.

Tools named in the evidence
  • ChatGPT