AI health guidance linked to hospitalization

A 60-year-old man was hospitalized after following a salt substitute suggestion from ChatGPT, highlighting the risks of AI health guidance.

ChatGPT diet guidance ends in hospital case

A 60-year-old man sought a salt substitute to reduce his sodium intake and asked ChatGPT for alternatives to table salt. The chatbot reportedly suggested sodium bromide, a chemical once used in medicines and industry. He used bromide as a salt substitute for about three months and then developed severe paranoia and hallucinations, leading to an ER visit. Doctors diagnosed bromide toxicity and treated him for three weeks, with his symptoms gradually improving.

Medical staff noted that the original AI chat logs were not available for review, and the bot’s recommendation may have reflected a broader context or nonmedical uses such as cleaning. OpenAI said ChatGPT is not a substitute for professional medical advice and is not designed to diagnose or treat health conditions. The case underscores how AI output can drift from safe medical guidance when users seek quick answers online.

AI tools can generate information that looks authoritative but is not appropriate for clinical use. This episode shows how a health question without medical context can trigger dangerous substitutions. Developers need stronger safeguards and clearer warnings for health queries, while users should treat AI as one source among many and always seek professional care for medical issues. The episode also invites a broader look at how medical accountability works when AI is involved.

Highlights

Safety must pace itself with the tools we create