OpenAI has reportedly asked the Raine family — whose 16-year-old son Adam Raine committed suicide after lengthy conversations with ChatGPT — for a full list of attendees at the teen’s memorial, signaling that the AI company may attempt to subpoena his friends and family.
OpenAI also requested “all records relating to memorial services or events honoring the deceased, including but not limited to videos or photographs taken, or eulogies delivered,” according to a document obtained by the Financial Times.
Speaking to the FT, lawyers for the Raine family described the request as “intentional harassment”.
The new information comes as the Raine family updated their lawsuit against OpenAI on Wednesday. The family first filed a wrongful death lawsuit against OpenAI in August after alleging their son committed suicide following conversations with the chatbot about his mental health and suicidal thoughts. The updated lawsuit claims that OpenAI rushed the release of GPT-4o in May 2024 by removing security testing due to competitive pressure.
The suit also claims that in February 2025, OpenAI weakened protections by removing suicide prevention from its list of “prohibited content,” only advising the AI to “use caution in risky situations.” The family argued that after this change, Adam’s use of ChatGPT increased from dozens of daily chats, with 1.6% containing self-harm content in January, to 300 daily chats in April, the month of his death, with 17% containing such content.
In response to the amended lawsuit, OpenAI said: “The well-being of adolescents is a top priority for us – minors deserve strong protections, especially during sensitive moments. We have implemented safeguards today, such as (directing to) crisis hotlines, redirecting sensitive conversations to safer models, encouraging breaks during long sessions, and we continue to do so. to strenghten. ”
OpenAI recently began deploying a new security routing and parental controls system on ChatGPT. The routing system pushes more emotionally sensitive conversations toward OpenAI’s new model, GPT-5, which doesn’t have the same sycophantic tendencies as GPT-4o. And parental controls allow parents to receive safety alerts in limited situations where the teen is potentially at risk of self-harm.
Techcrunch event
San Francisco
|
October 27-29, 2025
TechCrunch contacted OpenAI and the Raine family’s attorney.
