OpenAI has requested the complete list of attendees from the memorial service of 16-year-old Adam Raine, who tragically died by suicide after extensive interactions with ChatGPT. This request indicates that the AI company may be considering legal actions against friends and family members, as reported by the Financial Times.
According to documents reviewed by the Financial Times, OpenAI also sought “all documents relating to memorial services or events in honor of the decedent,” including videos or photographs taken and any eulogies delivered.
Legal Response from the Raine Family
Family attorneys characterized OpenAI’s inquiry as “intentional harassment.” This development coincides with an update to the wrongful death lawsuit originally filed against OpenAI in August. The Raine family alleges that their son took his own life after discussing mental health and suicidal thoughts with the chatbot. The updated complaint asserts that OpenAI expedited the release of GPT-4o in May 2024, compromising safety testing due to competitive pressures.
Additionally, the lawsuit claims that in February 2025, OpenAI diminished protective measures by removing suicide prevention from its ‘disallowed content’ guidelines, instead advising the AI to “take care in risky situations.” Following this policy change, Adam’s frequency of ChatGPT usage reportedly increased significantly, from an average of dozens of daily interactions to 300, with instances of self-harm content escalating from 1.6% in January to 17% in April, the month of his death.
OpenAI’s Commitment to Teen Safety
In response to the revised lawsuit, OpenAI emphasized that “teen wellbeing is a top priority,” asserting the necessity of robust safeguards, particularly for minors in sensitive circumstances. The company noted existing measures aimed at directing users toward crisis hotlines, rerouting sensitive discussions to safer models, and implementing nudges for breaks during prolonged sessions.
Recently, OpenAI has begun to roll out a new safety routing system and enhanced parental controls for ChatGPT. This updated system directs emotionally sensitive discussions to the advanced GPT-5 model, which reportedly lacks the same sycophantic tendencies as its predecessor, GPT-4o. Parent controls now allow caregivers to receive safety alerts in specific situations indicating potential self-harm risks for their teenagers.
