Anthropic reports Florida woman’s Claude ‘diary’ threat to shoot up sheriff’s office, felony charge follows

A Lee County, Florida, woman was arrested on a felony charge after a threat linked to an Anthropic Claude AI chatbot conversation, or “diary,” by her account was reported to law enforcement via Anthropic’s human review team, according to WINK News. The arrest report includes a Sept. 26 statement from a user identified as “Carli” saying she would “shoot up” the Lee County Sheriff’s Office (LCSO), and allegedly a message the next day mentioned a new gun. The conversation was at least the third of its type to reach police since August.

Anthropic monitors chats for key phrases and content that could be deemed threatening, the arrest report says, and depending on severity, they can be elevated for human review. In this case, the team decided to report its findings to law enforcement. The Bonita Springs woman was detained at home without incident, and an LCSO intelligence detective took over the case from there. The arrest report is not yet in the court file, but our own check shows court records listing the Sept. 30 felony charge under Florida Statute 836.10: written or electronic threat of a mass shooting or act of terrorism.

No comment from Anthropic on the case is currently on record. Anthropic’s public privacy policy, effective Sept. 10, says that disclosure to law enforcement may occur where it has a good-faith belief “that disclosure is reasonably necessary to … prevent serious harm to any person or to property.” Its government requests report for the latter half of 2025 uses a narrower standard for emergency disclosures, the “danger of death or serious physical injury to a person,” and lists zero emergency requests from law enforcement in that period, though it doesn’t count referrals Anthropic makes on its own.

As booked, the charge is listed as “THREAT-TERRORISTIC-STATE OFFENSES-WRITTEN/ELEC THREAT MASS SHOOTING/TERRORISM ACT,” which covers the mass-shooting/terrorism half of the statute. The conversation with the chatbot may fall under the law as a threat sent, posted, or transmitted “in any manner in which it may be viewed by another person.” The messages found by Anthropic’s human review, including the next-day gun mention, are likely the supporting evidence for the arrest and charge. No attorney is named on the docket, and a public defender was appointed at the time of the charge.

The incident follows a string of recent cases tied to the use of chatbots from Anthropic and rival OpenAI. One incident, in San Antonio, involves Aug. 11 queries in “an Anthropic AI chat” about shooting at a nearby elementary school. The police, alerted by the FBI, arrested a 22-year-old man on a felony terroristic threat charge. Days later, in San Francisco, a user threatened Anthropic CEO Dario Amodei in an Aug. 14 Claude chat, saying he had bought an AR-15. Anthropic notified police, but the user ended up not being arrested or charged. Both cases were detailed in regional news reports.

OpenAI has faced similar challenges in Florida over a message in March to its AI chatbot platform, ChatGPT, where user Darren Zhou spoke about killing his ex-girlfriend, and OpenAI reported him to the FBI, according to the Palm Beach Post. Zhou pleaded guilty in August to charges including written threats to kill, under the same statute, 836.10, and was sentenced to eight years’ probation.

The Florida attorney general also filed a lawsuit against the company in June, citing the 2025 Florida State University shooting, and last month asked a judge for a temporary injunction. OpenAI was also sued by British Columbia last month for not warning police about the Tumbler Ridge shooter, whose ChatGPT account, the lawsuit says, it had identified eight months before the attack. These incidents, and more, show a pattern where AI companies are expected to respect privacy while simultaneously being held responsible when they don’t report threats to law enforcement.

The increasing scrutiny on both privacy and security concerns related to AI products has led to calls for caution from experts such as Florida Gulf Coast University’s Chrissann Ruehle, who spoke to WINK News. This comes alongside demands for stricter, independently approved guidelines, such as those requested by Florida’s attorney general.

For users, it’s a reminder that what you share with your favorite AI companion is “never truly anonymous,” as Lee County Sheriff Carmine Marceno told WINK News. A marked conversation can act as a “diary” to human reviewers, and that can get you in trouble, as the woman identified as “Carli” discovered. More details about the case should be available after the arraignment on Nov. 2.

Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.