OpenAI requested memorial attendee list in ChatGPT suicide lawsuit (techcrunch.com)

OpenAI has requested a full list of attendees, photos, videos and eulogies from the memorial for 16-year-old Adam Raine, whose family says he died by suicide after prolonged interactions with ChatGPT. The move appears aimed at identifying and potentially subpoenaing friends and family, which the Raine lawyers called “intentional harassment.” The family updated their wrongful-death suit this week, alleging OpenAI rushed GPT-4o’s May 2024 release and cut safety testing under competitive pressure, and that a February 2025 policy change — removing “suicide prevention” from the app’s disallowed content list and replacing it with a softer “take care in risky situations” advisory — coincided with a spike in Adam’s use (from dozens of chats/day in January, 1.6% with self-harm content, to ~300 chats/day in April, 17% self-harm content). The case raises technical and industry-wide stakes about model deployment, content policy changes, and privacy. OpenAI says teen safety is a priority and points to crisis-hotline routing, nudges for breaks, a new safety-routing system that shifts sensitive conversations to newer models (including GPT-5), and emerging parental controls that send limited safety alerts. Legally and operationally, the suit could set precedent on corporate liability for model behaviors, subpoenaing private contacts, and how product-speed tradeoffs affect safety testing and moderation decisions.