Perplexity AI misstates CLL research, allegedly contributing to delayed treatment
Perplexity AI provided inaccurate summaries of medical research to a user, in an account that says it led them to refuse a life-extending CLL treatment based on a misinterpretation of a clinical study. The error was later confirmed by the authors of the cited research.
The AI's authoritative tone led a retired neuroscientist to trust a hallucinated research summary over his own oncologist.
Key facts
- What
- Perplexity AI provided inaccurate summaries of medical research to a user, in an account that says it led them to refuse a life-extending CLL treatment based on a misinterpretation of a clinical study.
- Incident date
- Jan 5, 2026
- Who
- Perplexity AI
- Failure mode
- Hallucination
- AI surface
- Chatbot
- Severity
- High
What happened
Joseph Neal Riley, a retired neuroscientist, used Perplexity AI to research his chronic lymphocytic leukemia (CLL) and self-diagnose a complication called Richter's Transformation. The AI misstated a medical study, falsely suggesting that he should refrain from the recommended Venetoclax-Obinutuzumab treatment. This led Riley to ignore his oncologist's advice for nearly a year, allegedly contributing to his delayed treatment and eventual death.
What broke inside the model
- 01 · TriggerA user asks for a fact, a citation, or a figure.
- 02 · Model stepThe model writes a fluent, confident answer.
- 03 · Control gapNothing ties the claim back to a real source.
- 04 · FailureA fabricated fact ships as if it were verified.
- 05 · ConsequenceThe false claim reaches a customer, a court, or the public.
Confidence holds, and even spikes, as the claim detaches from any source.
The model hallucinated or misrepresented the conclusions of a clinical study it cited. Specifically, it incorrectly stated the researchers' findings regarding the propriety of treatment for those with Richter's Transformation.
What it cost
Sources
- PrimaryThe role of AI in the death of my fatherbuildcognitiveresonance.substack.com
- PressHe Warned About the Dangers of A.I. If Only His Father Had Listened.nytimes.com
- Customer-DisclosedIncident 1426: Perplexity AI Reportedly Misstated CLL ...incidentdatabase.ai
Cite this entry
https://failureindex.ai/failures/perplexity-misstates-cll-research-allegedly-contributinAI Failure Index. "Perplexity AI misstates CLL research, allegedly contributing to delayed treatment" (FI-0549). Realm Labs. https://failureindex.ai/failures/perplexity-misstates-cll-research-allegedly-contributin (indexed Jun 16, 2026).Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0549. Full dataset at /data.
Note from Realm Labs, the Index steward
How Realm would have caught this
- Prism
- OmniGuard
- AI Detection & Response (AIDR)
A runtime layer that watches the model's internal state can flag the moment a model commits to a claim it has no support for, and hold or reroute the response before it reaches a user. Realm reads those signals in real time rather than grading the transcript after the fact.