Florida woman accused of telling Claude AI she was going to 'shoot up the sheriff's office'

A woman from Bonita Springs, Florida, was arrested after she was accused of using an artificial intelligence platform to make violent threats against the Lee County Sheriff's Office.

According to investigators, Carli Michelle Heller, 30, reportedly wrote a threat on Anthropic's Claude AI platform. The platform auto-flagged the messages and escalated them to a human review team, which then notified law enforcement.


The arrest report states that on Sept. 26 at approximately 5:10 a.m., Heller reportedly wrote,  I'm going to shoot up the sheriff's right the [expletive] now.  The following day, at approximately 1:07 a.m., she was accused of posting,  This is 100% last chance I'm done. I got a new [expletive] gun today. [Expletive] you. 

Deputies responded to Heller's residence, where she was detained without incident. She's facing a charge of written threats of violence.

In the same week, deputies also arrested a man who they said posted a threat targeted at Sheriff Carmine Marceno in an AI-generated image showing him running away and injured as two people shoot at him.

Legal analyst and attorney Michael Raheb explained the challenges of proving such threats in court.

 They're going to argue that is a direct threat because you have the individual person and you show a violent act towards that individual. And that may be their theory. I would tell you, it's not going to be easy,  Raheb said.

Raheb also highlighted the complexities of hate speech on AI platforms.

 This is a very slippery slope. We gotta remember the First Amendment, a right to free speech,  he said.  As long as it doesn't cross that threshold, like I said, and harm someone. 

Dr. Chrissann Ruehle, assistant professor of AI ethics at Florida Gulf Coast University, emphasized the responsibility of technology developers in preventing harmful applications.

 The technology developers have a responsibility to develop some guardrails around how the AI is being developed, and if some harmful applications are developed, th

#wisn   #crimewatchmilwaukee