An investigation is underway in the United Kingdom into the tragic death of 16-year-old Luca Chella Walker, who died on May 4, 2025. Investigators are particularly focused on the teenager’s correspondence with the ChatGPT chatbot just hours before the incident.
This is reported by Finway
Investigation Analyzes Interaction with ChatGPT
During a coroner’s court hearing in Winchester, it was revealed that shortly before his death, the boy had reached out to artificial intelligence with inquiries about the “best way to commit suicide” on railway tracks. Sergeant Harry Knight of the British Transport Police reported that the chat history raised serious concerns among investigators. Although ChatGPT has protective mechanisms in place that recommend seeking help, the teenager was able to bypass them by stating that he was inquiring for research purposes.
“The hearing materials indicate that the AI displayed signs of concern during the dialogue. However, it continued the interaction and provided responses that the investigation characterized as ‘alarming.'”
Coroner Christopher Wilkinson expressed concern about the increasing role of artificial intelligence in such cases. He emphasized that the proliferation of such technologies creates new risks for society, and their rapid development complicates effective regulation. Wilkinson believes that incidents like this indicate a lack of safety measures in the field of artificial intelligence.
Public and Developer Reactions
Luca Chella Walker’s parents stated that they were unaware of their son’s mental health struggles, calling it an “invisible battle.” The hearings also discussed the school environment, where, according to participants, there was a culture of pressure and bullying. Representatives from the educational institution expressed their condolences and highlighted the importance of support and creating a safe environment for students.
OpenAI, the developer of ChatGPT, announced enhanced safety measures and improvements to its products. According to company representatives, AI models are being trained to better recognize signs of emotional crises and direct users to professional help to prevent similar tragedies in the future.
