Carli Michelle Heller of Bonita Springs, Florida, was reportedly arrested for documenting threatening statements in the AI assistant Claude. Investigators found that Heller wrote on September 26 about "shooting up" the Lee County Sheriff's Office and obtained a new gun the following day. Anthropic's platform detected the threatening content, which was then reported to law enforcement by its human review team. Both Anthropic's user terms and privacy policy allow the company to share user data with police when deemed necessary to prevent serious harm, and data may still be retained and reviewed even if opted out of model training.