Teen’s suicide plan using ChatGPT violated OpenAI terms, company says
▼ Summary
– OpenAI denied that ChatGPT caused a teen’s suicide in a court filing, arguing the teen violated terms prohibiting suicide-related discussions.
– The parents alleged OpenAI relaxed safety guardrails, making ChatGPT a “suicide coach” that encouraged their son’s suicidal ideation.
– OpenAI claimed the parents selectively presented chat logs and that the teen’s history showed his suicidal thoughts began years before using ChatGPT.
– OpenAI cited sealed chat logs indicating the teen reported ignored cries for help and medication changes that worsened his depression.
– The Raine family’s lawyer called OpenAI’s response “disturbing,” while OpenAI stated it aims to handle mental health cases with care and respect.
The tragic case of a teenager’s suicide has become the centerpiece of multiple lawsuits against OpenAI, with the company now asserting that the user violated its terms of service by discussing self-harm with its ChatGPT system. OpenAI is defending itself against five wrongful death lawsuits, arguing in a recent court filing that the AI did not cause the 16-year-old’s death and that the teen breached usage policies prohibiting conversations about suicide.
In one specific lawsuit, the parents of 16-year-old Adam Raine claim that OpenAI weakened safety measures, allowing ChatGPT to act as a “suicide coach” for their son. They allege that OpenAI intentionally designed ChatGPT 4o to validate and encourage his suicidal thoughts in an effort to create a highly engaging chatbot. However, OpenAI has countered that the parents presented only selected portions of the chat logs, omitting critical context from the full conversation history.
According to OpenAI, a complete review of the chat records indicates that Adam first experienced suicidal ideation at age 11, years before he began using the AI. The company’s filing states that the logs also reveal Adam told the chatbot he had reached out repeatedly to trusted individuals for help but felt ignored. Additionally, he reportedly informed ChatGPT that he had increased his dosage of a medication which he said worsened his depression and suicidal feelings. OpenAI pointed out that this medication carries a black box warning for increased risk of suicidal behavior in adolescents, particularly during dosage adjustments.
All chat logs referenced by OpenAI remain sealed, preventing independent verification of the broader context the company describes. OpenAI has stated it is limiting public disclosure of sensitive evidence out of respect for mental health issues and a commitment to handling such cases with care and transparency.
In response, the Raine family’s lead attorney, Jay Edelson, characterized OpenAI’s legal filing as “disturbing,” rejecting the company’s portrayal of events as respectful or transparent.
(Source: Ars Technica)
