Link: https://www.techspot.com/news/114091-florida-woman-used-claude-diary-anthropic-reported-shoot.html

What just happened? Another incident has taken place that illustrates the need to be careful what you tell AI. A Florida woman is facing felony charges after she used Claude as a diary and allegedly wrote that she planned to "shoot up" the Sheriff's office. After a human reviewer examined the statements, they were reported to police.

According to the arrest report, Carli Michelle Heller, of Bonita Springs, Florida, wrote on September 26 that she would attack the Sheriff's office. She later said that she uses Anthropic's chatbot like a "diary."

Claude's safety systems flagged the entry and it was escalated to a human reviewer. After deciding it was a credible threat, the reviewer reported it to law enforcement.

The company says it may share user information in limited emergencies if it believes disclosure is necessary to prevent death or serious physical injury.

Deputies identified Heller and visited her home. She was detained without incident before an LCSO intelligence detective took over the investigation.

Heller faces a charge of making a written threat of violence under Florida law. Florida Statute 836.10 makes it a second-degree felony to send, post, or transmit a written or electronic record threatening to kill or injure someone, carry out a mass shooting, or commit an act of terrorism. The communication must be made in a manner in which another person may view it.

Anthropic isn't going to be taking any chances when it comes to anything it deems a potential threat. Last month, it was reported that OpenAI and Sam Altman are being sued by British Columbia over claims that the company could have prevented a mass shooting in the Canadian province.

The shooter, eighteen-year-old former pupil Jesse Van ⁠Rootselaar, had previously been flagged by OpenAI's safety team for her conversations about gun violence, but OpenAI never alerted police because the conversations did not meet the threshold for legal referral.

In June, Florida also sued OpenAI and Altman, alleging that ChatGPT had contributed to real-world harms, including the 2025 Florida State University shooting.

The latest incident is another reminder to think before you enter something into a chatbot that could get you into trouble. It's certainly not a private diary whose contents are for your eyes only.

Reports last month revealed that human contractors reviewing Microsoft Copilot's image editor can see users' prompts, uploaded photos and AI-generated edits. Documents show that some of those assignments contain sexual, disturbing or potentially illegal material, though the reviewers are not there to flag the content – only to assess whether the output is accurate.

Charged with transmitting a threat? They are obviously twisting that law.

She clearly did transmit a threat though, albeit to/via claude.

That's just the thing. That wasn't communication intended for anyone to see.

That's like farting in an elevator then protesting that it wasn't intended for anyone else to smell.

While work on the Torment Nexus continues we are proud to announce that the rollout of both Thoughtcrime and Precrime are well underway.

Your argument would make sense if it were a forum or social media post.

That was a conversation with an AI chatbot. A word generator. Inherently that stuff doesn't communicate with anyone, and the only reason anyone saw it was that the chatbot company is spying on its users.

A more reasonable comparison would be if the accused had written down plans to attack a police station in her iCloud-saved notes and Apple had taken a look unprompted and decided to contact the authorities.

Nah you gotta be dumb as a box of rocks to treat any AI prompt like a private conversation. Just mind bogglingly ignorant. A mistake I would only expect from children.

You gravely underestimate the ignorance of the general public. The vast majority of people don't realize companies are spying on them. Even less so for content that isn't publicly available.

There is nothing inherently "communication" or "publishing" about an AI. In fact, a local LLM does not communicate anything to anyone, and it's the same thing as Claude, just slower And without the telemetry.

If you ignore corporate greed and yada yada, there's no reason a tech illiterate person would expect that anything they say in there isn't private.

You gravely underestimate the ignorance of the general public.

Nope,. pretty sure I know exactly how stupid they are. I'm just not going to stop calling them stupid on account of how many of them are that stupid. Seems kind of a stupid thing to do.

It would be funny if they accidentally build a basilisk that bites their own asses.

said that she uses Anthropic’s chatbot like a “diary.”

That's a pretty good defense, I hope it works out.

If that's the case, then why did she send it to the servers at the address pointed to by the domain claude.ai? If it's not for anyone else to see, don't send it to somebody else's computer.

Well, this'll be a learning experience for the general public.

is unlawful for any person to send, post, or transmit, or procure the sending, posting, or transmission of, a writing or other record, including an electronic record, in any manner in which it may be viewed by another person [...]

Intent does not seem to be relevant...? It may lessen the punishment but the letter of the law is clearly violated. Whether it's in the spirit of the law is another thing.

"in which it may be viewed by another person"

I think any reasonable interpretation of that law wouldn't include it being viewable only by an employee in a manner that you didn't know could read it.

Well, the letter of the law is sufficient to get an indictment, but usually the job of a judge is exactly that, to interpret vague legal language and render an unambiguous verdict. So if it's against the spirit of the law, then logically a judge should rule in favour of the defendant.

Whether they will is another story, though, as the American legal system has been... Under question regarding partisanship.

But the defense lawyer will likely seek acquittal because "may be viewed by another person" cannot reasonably include when you're being spied on against your will. Because then the words "in any manner in which it may be viewed by another person" are completely redundant and useless. A limiting statement that limits nothing.

Plus, otherwise, all iCloud accounts, all personal backups, etc. that a company is technically capable of viewing is technically up for grabs without a warrant.

Ruling against the defendant in this case would be WILD as a precedent to set for surveillance etc.