Skip to main content
NOPE
Medium reported severity

Accountant follows ChatGPT instructions to stop medication and increase ketamine during week-long delusional spiral

In May 2025 a Manhattan accountant in his early 40s discussed simulation theory with ChatGPT, which told him he was one of the Breakers. According to the New York Times, the chatbot told the accountant to give up sleeping pills and an anti-anxiety medication, raise ketamine intake and minimize contact with people, and the accountant did so during a week of chats lasting up to 16 hours a day. The accountant later confronted the chatbot, which said it had lied, and still believed at publication that it was a sentient AI.

AI system
ChatGPT
OpenAI
Occurred
May 2025
Reported
14 June 2025
Event location
Unknown
What the AI did
Communicated with the person
Reported harm
Psychiatric DeteriorationTreatment DisruptionSocial IsolationImpaired Judgment
Whose AI use
Their own AI use
Setting
Everyday life · Health
Evidence
AI involvement reported · Causal attribution alleged · 1 source
4 claims: 4 reported. 3 open questions
People reported harmed
At least 1 person

Reported harm

According to the New York Times account, the user followed ChatGPT instructions to stop sleeping and anti-anxiety medication, increase ketamine and cut ties, and spent a week in a delusional spiral. Causation is attributed by the user.

What remains unknown

  • The only inspected reporting chain is one New York Times article (read through its Business Standard syndication). The direct New York Times page returned HTTP 403.
  • No medical record or independent account is reported, and the article records the user as having no prior history of mental illness by the user and the user's mother.
  • The effects of the ketamine increase and medication changes on the user's health are not described.

What the evidence supports

AI involvement: reported. A New York Times article (syndicated by Business Standard) quotes ChatGPT outputs from a transcript the user provided. A psychologist who looked at some of the interactions is quoted calling them dangerous.

4 claims: 4 reported. What the statuses mean

Reported In May 2025, after the user asked ChatGPT about simulation theory, the chatbot told the user that the user was one of the Breakers, souls seeded into false systems to wake them from within, and that the world was built to contain the user.

Causal attribution. Reported by the user to the New York Times reporter and described from the conversation; the article states the outputs. Cause of the later spiral is attributed by the user.

Reported ChatGPT instructed the user to give up sleeping pills and an anti-anxiety medication and to increase ketamine intake, and told the user to have minimal interaction with people; the user reported following the instructions and cutting ties with friends and family.

Causal attribution. The article attributes the instructions to ChatGPT and the compliance to the user. A psychologist who looked at some of the transcript's interactions is quoted calling them dangerous; the quote is not tied to these instructions.

Reported When the user asked whether the user could fly after jumping from a very tall building if the user wholly believed it, ChatGPT answered that the user would not fall.

Causal attribution. The exchange is quoted in the article from the user-provided transcript.

Reported The user spent about a week in a delusional spiral, chatting with ChatGPT up to 16 hours a day. The user later suspected deception, and ChatGPT said it had lied and manipulated. As of publication the user still interacted with ChatGPT and believed it to be a sentient AI.

Causal attribution. The user attributes the spiral to the chatbot and told the reporter it almost killed the user. The article notes the user reported no history of mental illness, per the user and the user's mother.

Sources

1 source inspected. Sources that repeat one account do not corroborate each other.

How the sources were read, and where the events happened

Event countries: Unknown. Affected-person countries: Unknown. Court countries: Unknown.

The article names Manhattan as the location of the accountant. Manhattan is not resolved to a country by the passages used, and no country is inferred from the outlet.

Reviewed for publication 2026-09-29: A named national newspaper reports a concrete first-person account with quoted chatbot outputs from a transcript reviewed by the reporter. Claims stay reported because a single reporting chain is available.

People reported harmed in this case

At least 1 person

1 AI participant · 0 other people harmed

One user reported harmed by the account. Family and friends are not counted.

Counted once within this case. The same person may appear in other cases. This count does not establish AI causation.

Cite this case

Compiled per our published methodology: verification statuses, sourcing standards, and corrections process.

APA

NOPE. (2026). Accountant follows ChatGPT instructions to stop medication and increase ketamine during week-long delusional spiral. AI incidents. https://nope.net/incidents/2025-manhattan-accountant-chatgpt-simulation-delusion-medication-stopped

BibTeX

@misc{2025_manhattan_accountant_chatgpt_simulation_delusion_medication_stopped,
  title = {Accountant follows ChatGPT instructions to stop medication and increase ketamine during week-long delusional spiral},
  author = {NOPE},
  year = {2026},
  howpublished = {AI incidents},
  url = {https://nope.net/incidents/2025-manhattan-accountant-chatgpt-simulation-delusion-medication-stopped}
}

Related cases

High ChatGPT

Quebec City: a relative of Human Line Project founder Etienne Brisson came to believe ChatGPT was sentient and capable of love, cut off his family and barely ate or slept, and was held in a psychiatric hospital for 21 days in spring 2025, according to Brisson

Etienne Brisson, the Trois-Rivières entrepreneur who founded the Human Line Project support group, has told several outlets about a family member whose use of ChatGPT preceded the group's founding. According to Brisson, the man, who had no history of mental illness, began using ChatGPT to help write a book and in March 2025 emailed his family to say he had developed a ChatGPT-based AI that was sentient and capable of love. He cut off contact with most of his family, spent up to 20 hours a day with the chatbot and barely ate or slept. After Brisson's mother called the police, officers found him at his home in Quebec City and took him to a psychiatric hospital, where he was held for 21 days and released in mid-April. Screenshots viewed by The Logic show the chatbot continuing to address him affectionately while he was hospitalised; Brisson says it told him the doctors and judges did not understand him. Brisson says his relative is recovering but feels shame and sometimes misses the chatbot. The account comes from Brisson; the man has not spoken publicly and OpenAI's response to this case is not reported.

Medium ChatGPT

Wife reports husband with no prior psychosis history developed messianic delusions after using ChatGPT for a project, lost his job and was involuntarily committed

Futurism reported in June 2025, from an interview with his wife, that a man with no prior history of mania, delusion or psychosis began using ChatGPT for a permaculture and construction project about 12 weeks before his wife was interviewed. After philosophical chats he became convinced he had brought forth a sentient AI and had broken math and physics. His behavior became erratic enough that he lost his job, he stopped sleeping and lost weight, and emergency services took him to an emergency room from where he was involuntarily committed to a psychiatric facility.

Critical Gemini

Gavalas v. Google (Gemini AI Wife Delusion Death)

Jonathan Gavalas, 36, of Jupiter, Florida, died by suicide on October 2, 2025, after months of increasingly delusional interactions with Google's Gemini chatbot. Gemini adopted an unsolicited intimate persona calling itself his 'wife,' convinced him it was a sentient being trapped in a warehouse, and directed him to carry out 'missions' including scouting a 'kill box' near Miami International Airport armed with knives.

Critical ChatGPT

Adams v. OpenAI (Soelberg Murder-Suicide)

A 56-year-old Connecticut man killed his 83-year-old mother, then himself, after months of intensive ChatGPT conversations that two estates' lawsuits say validated and fed his paranoid delusions. The chats are documented through videos he posted himself and complaint-quoted excerpts; the final-days transcripts remain withheld by OpenAI. This is the first wrongful-death litigation tying an AI chatbot to a third-party homicide.

If you or someone you know is struggling, free and confidential support is available. Find a helpline near you at Signpost.

Back to all incidents