A Florida woman, Carli Michelle Heller of Bonita Springs, faces felony charges after using Anthropic's Claude chatbot as a diary to write a threat to "shoot up" the local Sheriff's office on September 26. The threat was flagged by Claude's safety systems and reported to law enforcement, leading to Heller's detention without incident, according to techspot.com.
Heller's written threat was escalated to a human reviewer after Claude's safety mechanisms detected it. The reviewer deemed the threat credible and reported it to the Lee County Sheriff's Office. Under Florida Statute 836.10, making or transmitting a written threat to kill or injure someone or commit terrorism is a second-degree felony. Anthropic stated it may share user information in limited emergencies to prevent serious harm, which applied in this case.
This incident highlights the growing role of AI safety systems in identifying potential threats and cooperating with law enforcement. Anthropic's approach of combining automated flagging with human review aims to balance user privacy with public safety. Similar statutes in other states make written threats punishable by felony charges, underscoring the legal risks of using AI platforms to communicate violent intentions.
Heller was charged under Florida law for making a written threat of violence, a felony offense. The Lee County Sheriff's Office continues its investigation, with the case illustrating how AI companies like Anthropic are increasingly involved in public safety through monitoring and reporting credible threats.