Anthropic says minors have used its Claude AI

In its IPO filing submitted to the U.S. Securities and Exchange Commission, Anthropic disclosed that minors had used Claude, even though the rules prohibit users under 18 from using it; people at risk of suicide or self-harm had also used its AI systems under safeguards. The filing acknowledges that the long-term effects of regular interaction with AI on mental health, cognitive development and social behavior are not yet fully understood, and mentions that other companies' AI systems have been alleged to be linked to suicide, violence and serious self-harm, stating that similar incidents involving its own models have posed a material risk to its reputation, but it does not state that Claude caused such incidents, nor does it provide details. Anthropic says its safety systems screen conversations and monitor for content that encourages harmful thoughts, and can adjust responses in real time, warn users, or in severe cases close accounts; however, because a zero data retention agreement does not retain the conversations themselves, the company cannot look back at past conversations for most traffic to investigate incidents. The filing does not report the frequency of such use, the number of minors involved, or the actual performance of the safeguards.