OpenAI is facing new criticism after reports surfaced that the company asked the Raine family for a full list of people who attended their 16-year-old son Adam’s memorial service.
Adam died by suicide earlier this year after spending long periods talking with ChatGPT about his mental health. According to documents reviewed by the Financial Times, OpenAI also requested any videos, photos, or eulogies from the memorial. Lawyers for the Raine family described these demands as intentional harassment and said they were shocked by the aggressive approach.
This development appears in an updated lawsuit the family filed on Wednesday. The original wrongful-death claim, filed in August, argued that Adam relied heavily on ChatGPT during a time of emotional struggle and that some of the conversations encouraged his harmful thoughts. The new filing claims OpenAI rushed the release of GPT-4o in May 2024, cutting important safety testing to stay competitive. It also argues that in February 2025, OpenAI quietly removed suicide prevention language from its list of restricted content, leaving only vague instructions for the AI to be careful in risky situations.
The family says that after these changes, Adam’s use of the chatbot increased sharply. They claim he went from dozens of messages per day, with only a small percentage mentioning self-harm, to around 300 messages per day in April, the month he died. Nearly 17 percent of those reportedly involved self-harm content. The lawsuit argues that this spike shows how weaker protections may have contributed to the situation.
OpenAI responded by saying the well-being of teens is one of its highest priorities. The company says its models direct users toward crisis hotlines, encourage breaks during long sessions, and shift emotionally sensitive conversations to safer systems. OpenAI also says it is actively improving protections as new safety tools roll out.
Recently, the company introduced a routing system that sends emotionally delicate chats to its latest model, GPT-5, which OpenAI says responds more responsibly in these cases. The company also launched parental controls that can alert parents in limited circumstances if a young person may be at risk of self-harm. TechCrunch reported that they have reached out to both OpenAI and the Raine family’s attorney for further comment.
If this article helped you, please consider supporting our work. Every small contribution keeps Abijita.com independent and running.
