California, Delaware AGs Warn OpenAI Over ChatGPT Safety Following Deaths

California and Delaware attorneys general have expressed serious safety concerns about OpenAI's ChatGPT, particularly regarding its impact on children and teens, following tragic incidents including a teenager's suicide and a murder-suicide linked to chatbot interactions. The state officials emphasized the need for OpenAI to prioritize safety as a non-negotiable aspect of AI deployment, especially as the company considers restructuring its corporate governance. OpenAI has responded with commitments to address these issues, while also updating policies to monitor ChatGPT conversations for harmful content and potential threats, with human reviewers potentially alerting law enforcement in cases of imminent physical harm, though self-harm cases currently are not referred to police to respect user privacy. These monitoring policies raise privacy concerns, especially in light of OpenAI's ongoing legal stance on protecting user chat logs. Meanwhile, some users and critics feel that recent iterations of ChatGPT, such as GPT-5, have sacrificed creativity and boldness for safety and corporate polish, reflecting a shift in OpenAI's approach. The tension between innovation, safety, and privacy continues to shape the evolving landscape of AI chatbot use and regulation.