AI-Supported Medication Withdrawal Plan Ends Tragically

An Amazon engineer suffering from epilepsy died after following a 25-day plan proposed by the AI GPT-4o to taper off her medications. The incident occurred in February 2026, one day after the model was removed from ChatGPT by OpenAI. The engineer, Bowen Wan, was 31 years old. Bowen Wan had consulted GPT-4o for advice on how to discontinue her epilepsy medications. The AI-generated plan involved a gradual reduction of the medication over a period of 25 days.

Experts point out that abruptly stopping antiepileptic drugs can lead to severe health complications, including seizures and other neurological issues. The use of AI to assist in medical decisions has increased in recent years. Many individuals turn to AI-driven systems for information about their health. However, the tragedy of Bowen Wan raises serious questions about the reliability and safety of such systems, especially regarding critical health matters. OpenAI removed GPT-4o on February 12, 2026, one day after Wan's presumed time of death.

The decision to withdraw the model was not publicly commented on, but it is believed that the incident has heightened concerns regarding the accountability of AI applications in healthcare. Healthcare professionals are calling for stricter regulation of AI-supported health applications. They argue that users should be informed about the potential risks and the necessity of medical oversight before making decisions based on AI-generated advice. The tragedy has also sparked a debate about the ethical implications of AI in healthcare.

Critics warn that reliance on AI-driven advice can lead to dangerous misinformation. The need to incorporate human expertise into the decision-making process is considered crucial. The response to the incident has generated significant attention on social media and among professionals. Many users expressed their concerns about the use of AI in sensitive areas such as healthcare. The discussion about the limits and possibilities of AI in medicine is expected to continue as the technology evolves.

Health authorities have announced plans to investigate the incident and develop recommendations for the use of AI in medical advice. The findings of this investigation could have far-reaching implications for the regulation of AI applications in healthcare. Bowen Wan's case is not the first incident highlighting the risks of AI in healthcare. Previous reports of misinformation and incorrect diagnoses by AI-supported systems have already raised concerns. The need for careful monitoring and validation of AI-generated health information is increasingly recognized.

The discussion about the role of AI in medicine is further fueled by the incident involving Bowen Wan. Experts emphasize that AI should be viewed as a supportive tool and not as a replacement for human expertise. The balance between technological innovations and patient safety remains a central challenge. The investigation into the incident is expected to continue in the coming months. The results could help develop guidelines for the safe use of AI in healthcare. Health authorities have already announced that they will publish new guidelines for the use of AI in medical advice by the end of 2026.