Man Suffers Psychosis After Following AI advice to Replace Chloride with Bromide
BOSTON, MA – A man experienced a severe psychotic episode and required hospitalization after intentionally consuming sodium bromide based on details provided by the AI chatbot ChatGPT, according to a recently published case report. the incident highlights the potential dangers of relying on artificial intelligence for medical or health advice.
The case, detailed by doctors, unfolded when the patient sought a solution to concerns about sodium chloride (table salt) intake. Finding limited information on reducing sodium, he turned to ChatGPT seeking a way to eliminate chloride from his diet altogether. The AI reportedly suggested bromide as a safe substitute. He then began purchasing and consuming sodium bromide online.
Within three months, the man developed a full-blown psychotic episode, leading to an attempted escape from a medical facility and subsequent placement on an involuntary psychiatric hold due to “grave disability.” Doctors quickly stabilized him with intravenous fluids and antipsychotic medication, and began to suspect bromism – bromide toxicity – as the cause.
Once able to communicate, the patient confirmed he had been following ChatGPT’s advice. While the specific interaction with the AI (likely ChatGPT 3.5 or 4.0) remains unknown,testing by the doctors replicated the response: when asked what could replace chloride,ChatGPT included bromide in its suggestions.
Crucially,the AI’s response,while acknowledging the importance of context,failed to warn of the significant health risks associated with bromide consumption,nor did it inquire about the user’s reasoning for the question. the doctors noted the AI’s suggestion may have stemmed from unrelated applications of bromide,such as in cleaning products.
The man ultimately recovered and was discharged from the hospital after three weeks, remaining stable at a two-week follow-up.
The Growing Risks of AI-Driven Self-Treatment
This case serves as a stark warning about the limitations and potential dangers of relying on AI for health information. While AI tools like ChatGPT can offer access to information and potentially bridge the gap between scientific research and the public, they are prone to providing “decontextualized information” without the critical nuance and safety considerations of a qualified medical professional.The incident underscores several key concerns:
Lack of Contextual Understanding: AI chatbots ofen lack the ability to understand the full context of a user’s query, leading to potentially harmful recommendations.
Absence of Critical Inquiry: A human doctor would routinely ask clarifying questions to understand why a patient is seeking such information, a step the AI failed to take.
Failure to Warn of Risks: The AI did not provide any warnings about the dangers of bromide toxicity, a potentially life-threatening condition.
the Illusion of Authority: The AI’s response may have lent a false sense of authority to the suggestion, encouraging the man to proceed despite the inherent risks.
This case isn’t an isolated incident. As AI becomes increasingly integrated into daily life, the potential for misinformation and self-treatment based on flawed AI advice is growing. Experts emphasize the importance of verifying information obtained from AI with trusted sources, and, most importantly, consulting with a qualified healthcare professional for any health concerns.
