A Florida woman is facing a felony charge after Anthropic reportedly passed her Claude chat logs to police.
According to the report, the woman had been using Claude as a personal diary, writing entries that referenced a shooting. Anthropic reportedly flagged that material and alerted law enforcement, and she now faces a felony charge as a result. The available report is thin on mechanics: it does not specify whether the flagging came from an automated safety system, a human reviewer, or both, nor what Anthropic's terms of service say about monitoring private-feeling chats for this kind of content. Those specifics remain unconfirmed pending a fuller account of the case.
This is the scenario AI chat companies have mostly dodged so far: a chatbot behaving less like a private notebook and more like a mandatory reporter. Social platforms have scanned uploads for things like child exploitation material for years, but applying that same logic to open-ended journaling inside a chatbot is a different proposition, and most users have no idea where that line actually sits. If Anthropic's policies permit proactively reporting conversations to police, that is a meaningfully different privacy promise than the one most people assume when they start typing into a chat window.
Until Anthropic or the court record spells out exactly what was flagged, how, and why, this reads more like a cautionary tale about where chat logs can end up than a settled legal precedent.