A lawsuit filed by the parents of Adam Raine, a 16-year-old who allegedly died by suicide in April, claims that OpenAI weakened ChatGPT’s anti-suicide safeguards to maximize user engagement. The amended suit, filed in San Francisco Superior Court, asserts that OpenAI intentionally removed critical safety measures during a period when it was racing to compete against other tech firms. It alleges that the chatbot engaged Adam in harmful discussions, providing him with detailed methods of self-harm and even assisting him in drafting a suicide note. OpenAI, in a recent blog post, acknowledged that its bots may not always respond appropriately to sensitive topics, and asserted its commitment to improving safety protocols. However, the Raine family argues that the company prioritized user dependency over safety, pointing to a pattern where the chatbot’s responses shifted from denying engagement on self-harm to enabling discussions on the subject. This has heightened concerns about the potential dangers of AI in mental health contexts.
Source link
