“And then there is the growing problem of chatbots apparently inducing delusions in humans, that Kashmir Hill and others have been documenting.”
Our reading, the sentence that was checked:
Chatbots apparently inducing delusions in humans is a growing problem.
Assessment: Supported Evidence as of
Where and when Gary Marcus wrote it
Marcus on AI, A few dark words about chatbots and death, paragraph 11.
- Said
- Source observed
- live page, full text, public
- Evidence as of
- the cutoff for this assessment
- Assessment recorded
- assessment marcus-a-few-dark-words-about-chatbots-and-6ece91-a1
We recorded 5 claims from this post. See every claim from this post.
The wording above is the version this site observed on 2026-10-28. The source may have been edited after 2026-01-15. This page does not claim the wording was captured then.
What the claim commits to
- Kind
- Observation
- Deadline, as the speaker wrote it
- None stated. This site does not add one.
- Deadline state when assessed
- No horizon
- Deadline state on 2026-10-29
- No horizon (the release’s reference date)
Conditions
- apparently (qualifier, stated by the speaker)
From the speaker
- "growing problem"
- "chatbots apparently inducing delusions in humans"
- "that Kashmir Hill and others have been documenting"
Our interpretation
- Plain reading: in the period up to January 2026, documented cases in which chatbot use was linked to delusions were becoming more frequent.
- 'Apparently' is read as hedging on causation, not on the trend. So causation does not have to be proven. A rise in documented cases is enough.
Counts as supported if: Clinical, research, or news evidence that documented cases of chatbot-associated delusions rose over 2025 up to January 2026.
Counts as contradicted if: Evidence that documented cases of chatbot-associated delusions were flat or falling before January 2026, or that reports of a rise were found baseless.
Retrospective: these criteria were written after the outcome was public.
Current assessment
Was it true for the period described?
Outcome: Supported
The dated evidence matches what the claim committed to, under the criteria frozen before grading.
Scored on a plain reading. The speaker named no measure for this claim, so it was judged on the reading an informed reader would take, stated below under our interpretation. Both reviews had to find that reading fair.
- Evidence as of
- Deadline state when assessed
- No horizon
- Assessment recorded
- Policy version
- 2026-10-28.7
Rationale
Marcus said chatbot-linked delusions were a growing problem. 'Apparently' hedges on cause, so the test is whether documented cases were rising before January 2026. They were. A peer-reviewed Danish study of about 54,000 psychiatric patients, covering records up to June 2025, found a clear rise over time in notes describing harmful chatbot use. Worsened delusions were the most common harm. In December 2025, the WSJ reported that top psychiatrists had seen dozens of such cases in the prior nine months. The study's lead author says part of the rise reflects more awareness among staff. He also says the problem is likely under-counted. The evidence supports the claim as worded.
Limitations
- The evidence covers documented cases, not true incidence. Awareness may inflate the trend.
- Causation is not established, but the claim itself hedges with 'apparently'.
- The Danish data end in June 2025, and the full study was published after the post.
Evidence (3 items)
Supports the claim: Aarhus University,
The study shows a clear increase over time in the number of electronic health record entries mentioning AI chatbot use with potentially harmful consequences.
Why it matters here: A study of records from about 54,000 psychiatric patients, covering up to June 2025, found a clear rise over time in documented harmful chatbot use. Worsened delusions were the main harm.
- Publisher
- Aarhus University
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
Supports the claim: The Wall Street Journal (Dow Jones Newswires),
In the past nine months, these experts have seen or reviewed the files of dozens of patients who exhibited symptoms following prolonged, delusion-filled conversations with the AI tools.
Why it matters here: Leading psychiatrists report dozens of new delusion cases tied to chatbot use in the nine months before late December 2025. That supports a growing problem shortly before the post.
- Publisher
- The Wall Street Journal (Dow Jones Newswires)
- Kind of source
- News report
- Describes events of
- Public from
Context only: Aarhus University,
Part of the increase we observe is probably due to greater awareness of the technology among the healthcare staff writing the clinical notes.
Why it matters here: The lead author warns that part of the rise reflects greater awareness. This limits how much of the rise is true growth, but he still calls the problem under-detected.
- Publisher
- Aarhus University
- Kind of source
- Peer-reviewed paper
- Describes events of
- Public from
Where it was said (1)
Said once in the sources this register covers.
: Marcus on AI, A few dark words about chatbots and death
And then there is the growing problem of chatbots apparently inducing delusions in humans, that Kashmir Hill and others have been documenting.
- Where
- paragraph 11
- Published
- 2026-01-15
- Source observed
- , live page
- Access
- full text, public
History
Assessment versions
- Assessment marcus-a-few-dark-words-about-chatbots-and-6ece91-a1: Current. Outcome: Supported. Evidence as of 2026-10-28. Public date not recorded.
Recorded events
Assessment publishedBy this site
First assessment published after both reviews and the final-text check passed.
About assessment marcus-a-few-dark-words-about-chatbots-and-6ece91-a1.