Florida woman used Claude as a diary, then Anthropic reported an entry to police
According to the arrest report, Carli Michelle Heller, of Bonita Springs, Florida, wrote on September 26 that she would attack the Sheriff's office. She later said that she uses Anthropic's chatbot like a "diary." Read Entire Article
In Bonita Springs, Florida, Carli Michelle Heller, a woman, faces felony charges after she utilized Claude, an AI chatbot, as a diary. On September 26, she reportedly wrote on the platform that she intended to attack the Sheriff's office. Upon review, her statement was reported to local law enforcement. Claude's safety systems had initially flagged the entry, escalating it to a human reviewer. After determining the threat to be credible, the reviewer reported the matter to law enforcement.
Deputies identified Heller and visited her home. She was detained without incident, and an LCSO intelligence detective took over the investigation. Heller now faces a charge of making a written threat of violence under Florida law. According to Florida Statute 836.10, it is a second-degree felony to write, post, or transmit a communication threatening someone's life, infliction of injury, mass shooting, or act of terrorism, provided that the message can be seen by others.
Anthropic, the company behind Claude, takes no chances when it encounters potential threats. In a previous case, OpenAI and Sam Altman are facing legal action from British Columbia for alleged failure to prevent a mass shooting, despite warnings from OpenAI's safety team about the shooter's gun violence discussions. Similarly, Florida recently sued OpenAI and Altman for alleged contribution to real-world harms, including a 2025 Florida State University shooting.
This incident serves as a stark reminder that users should exercise caution when sharing information with AI chatbots. These platforms are not private diaries with content accessible only to the user. It is crucial to consider the potential consequences before entering anything into a chatbot, as reports suggest that human contractors reviewing Microsoft Copilot's image editor may see users' prompts, uploaded photos, and AI-generated edits, even if the reviewers are not there to flag the content but to assess the accuracy of the output.
Written by urgent.news from TechSpot's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
- Anthropic’s answer to Dots and Muse is already inside Claude thenewstack.io
- Sam Altman isn't happy that Anthropic asked religious scholars if Claude is conscious timesofindia.indiatimes.com
