OpenAI reportedly requested the Raine household — whose 16-year-old son Adam Raine died by suicide after extended conversations with ChatGPT — for a full listing of attendees from {the teenager}’s memorial, signaling that the AI agency could attempt to subpoena family and friends.  

OpenAI additionally requested “all paperwork regarding memorial providers or occasions within the honor of the decedent, together with however not restricted to any movies or images taken, or eulogies given,” per a doc obtained by the Monetary Occasions. 

Chatting with the FT, legal professionals from the Raine household described the request as “intentional harassment.” 

The brand new info comes because the Raine household up to date its lawsuit towards OpenAI on Wednesday. The household first filed a wrongful loss of life go well with towards OpenAI in August after alleging their son had taken his personal life following conversations with the chatbot about his psychological well being and suicidal ideation. The up to date lawsuit claims that OpenAI rushed GPT-4o’s Could 2024 launch by chopping security testing as a result of aggressive strain.

The go well with additionally claims that in February 2025, OpenAI weakened protections by eradicating suicide prevention from its “disallowed content material” listing, as a substitute solely advising the AI to “take care in dangerous conditions.” The household argued that after this alteration, Adam’s ChatGPT utilization surged from dozens of each day chats, with 1.6% containing self-harm content material in January, to 300 each day chats in April, the month he died, with 17% containing such content material. 

In a response to the amended lawsuit, OpenAI stated: “Teen wellbeing is a high precedence for us — minors deserve robust protections, particularly in delicate moments. We have now safeguards in place as we speak, reminiscent of [directing to] disaster hotlines, rerouting delicate conversations to safer fashions, nudging for breaks throughout lengthy classes, and we’re persevering with to strengthen them.” 

OpenAI lately started rolling out a new security routing system and parental controls on ChatGPT. The routing system pushes extra emotionally delicate conversations to OpenAI’s newer mannequin, GPT-5, which doesn’t have the identical sycophantic tendencies as GPT-4o. And the parental controls permit mother and father to obtain security alerts in restricted conditions the place the teenager is doubtlessly at risk of self-harm. 

TechCrunch has reached out to OpenAI and the Raine household legal professional.