Anthropic reported a Claude diary entry to police, and a Florida woman now faces a felony

A Florida woman is facing a felony charge after Anthropic reported one of her Claude conversations to the police. According to TechSpot, she had been using the chatbot as a personal diary, and an entry flagged by Claude's safety systems was escalated to a human reviewer who judged it a credible threat.

What happened

According to the arrest report described by TechSpot, Carli Michelle Heller of Bonita Springs, Florida, wrote in Claude on September 26 that she planned to attack the Sheriff's office. When questioned, she said she uses Anthropic's chatbot as a "diary."

Claude's automated safety systems flagged the entry and escalated it to a human reviewer. After that reviewer decided the statements amounted to a credible threat, the company passed the information to law enforcement. Anthropic says it may share user information in limited emergencies when it believes disclosure is necessary to prevent death or serious physical injury.

Deputies identified Heller, went to her home and detained her without incident before a sheriff's office intelligence detective took over the investigation.

The charge Heller faces

Heller is charged with making a written threat of violence. TechSpot outlines the relevant statute, Florida Statute 836.10, which makes it a second-degree felony to send, post or transmit a written or electronic record threatening to kill or injure someone, to carry out a mass shooting or to commit an act of terrorism. The law also requires that the communication be made in a way another person could view it — a condition that raises questions about how it applies to words typed into a private chat window.

AI providers are being pulled from both directions

The case lands amid litigation over when AI companies should involve police at all. TechSpot reports that British Columbia is suing OpenAI and Sam Altman, claiming the company could have prevented a mass shooting in the province. The shooter, 18-year-old former pupil Jesse Van Rootselaar, had earlier been flagged by OpenAI's safety team over conversations about gun violence, but OpenAI never alerted police because those conversations did not meet the threshold for a legal referral. Florida also sued OpenAI and Altman in June, alleging ChatGPT contributed to real-world harms including the 2025 Florida State University shooting.

The contrast is sharp: one company faces a lawsuit for not reporting flagged conversations, while another company's report feeds a felony charge. In both scenarios, the decision about whether a user's words cross the line rests with the provider's own reviewers, applying thresholds users never see.

Human review of user content is not unique to Anthropic. TechSpot also points to earlier reporting that contractors reviewing Microsoft Copilot's image editor can see users' prompts, uploaded photos and AI-generated edits — some of it sexual, disturbing or potentially illegal — even though those reviewers are hired to assess output accuracy rather than police the content.

Why it matters

The core lesson is that a chatbot is not a diary. Conversations with services like Claude pass through automated safety systems and, at times, human reviewers, and the provider reserves the right to hand the contents to police when it judges an emergency. For users, that means venting into a chat window can carry real legal exposure — in this instance, a second-degree felony charge. For the industry, it shows there is no comfortable default: stay quiet about flagged conversations and risk lawsuits, or report them and face hard questions about surveillance of intimate, diary-like exchanges. As AI journaling and companion apps become more common, the line between private reflection and reportable content is turning into a live legal question rather than a theoretical one.