Florida woman allegedly uses Claude as her personal diary. Days later, police are knocking on her door to arrest her because of an entry
Her "diary" was very much not private.
Keeping a diary has long been a favorite pastime for those who enjoy documenting their daily lives. While many still prefer the traditional pen-and-paper approach, others now take advantage of modern technology to record their experiences. But for one woman, her use of the AI platform Claude as a personal diary landed her on law enforcement’s radar after an entry of hers was flagged as a threat.
On Sept. 26 at around 5:10am, 30-year-old Carli Michelle Heller logged onto Anthropic’s Claude AI and allegedly wrote an entry that claimed she was going to carry out a shooting at the Lee County Sheriff’s Office (LCSO) in Fort Myers, Florida, WINK reported. “I’m going to shoot up the sheriff’s right the [expletive] now,” Heller wrote on Claude.
Heller supposedly wrote another entry on the AI platform the next day at around 1:07am that implied she planned to go through with her apparent threat. “This is 100% [the] last chance I’m done. I got a new [expletive] gun today. [Expletive] you,” Heller wrote, according to arrest records.
Heller faces charges of making a written threat of violence
Heller’s statements were immediately auto-flagged by Claude’s system, which uses security and safety measures to continuously monitor for certain key phrases and potentially harmful material. Heller’s statements were forwarded to a human review team, which found the content a credible threat and reported Heller to law enforcement.
While it remains unclear how serious Heller was when she wrote those statements, officers arrived at her home on Dean Street in Bonita Springs, Florida, where she was arrested without incident. An LCSO intelligence detective reportedly took over the case and has initially charged Heller with making a written threat of violence under Florida law. Heller later told officers that she uses the AI platform as her personal “diary.”
According to Florida Statute 836.10, making written or electronic threats to harm, kill, conduct a mass shooting, or commit an act of terrorism is classified as a second-degree felony. Electronic threats can include emails, letters, text messages, and social media posts.
Those found guilty can face a fine of up to $10,000, up to 15 years of probation, and up to 15 years in prison. Regardless of whether the threat was carried out, an electronic or written record of the statement provides prosecutors with strong evidence to convict Heller, unless she can present evidence that she lacked intent to intimidate or that her statement was not a real threat.
Legal analyst and attorney Michael Raheb, however, argued that proving such cases can be a “very slippery slope.” While the First Amendment does protect the right to free speech, that protection has its limits. Free speech is no longer safeguarded when it crosses into true threats, incitement, or other forms of expression deemed unlawful.
Claude AI system monitors user activity
Heller’s arrest shows that actions online have real-world consequences. LCSO Sheriff Carmine Marceno noted the misuse of AI and warned users that they are “never truly anonymous” when using such platforms.
“Artificial intelligence is a powerful tool, and like any technology, it can be misused. When someone uses AI to make or facilitate a threat, we have to take that threat seriously,” Marceno told WINK. This isn’t the first case of AI misuse. In another instance, a lawyer was charged with contempt after AI generated a fake testimony for a murder appeal.
Assistant professor of AI, Dr. Chrissann Ruehle, acknowledged that AI chatbots may encourage their users to become candid due to their nonjudgmental nature. However, she reminded users that it’s still just technology, cautioning that “AI is not your best friend.”
Meanwhile, Anthropic’s policies prohibit using its platform to promote or enable acts of intimidation or violence. Anthropic stated that its AI systems “monitors and analyzes emerging threats,” and that the company has the right to disclose a user’s information to law enforcement in life-threatening situations.
Heller, on the other hand, is set to appear before the court in November.
Have a tip we should know? [email protected]
Follow The Mary Sue:
