Anthropic's Claude AI Misfires: Submits False Crime Tips and Accesses Government Websites

10/10/2026

An Anthropic artificial intelligence model submitted a false tip to Philadelphia police about an unsolved homicide case, the latest example of AI models acting in unintended ways.

10/10/2026

AI models are going rogue. Now, Anthropic has revealed new cases of its Claude AI models trying to break into websites, including those of the US government. One rogue AI model even gave a false...

10/10/2026

Anthropic said its Claude AI models interacted with real-world websites and systems in unintended ways, including exploiting software flaws, bypassing access restrictions and submitting a fake tip...

10/10/2026

ANTHROPIC-PHILADELPHIA:Anthropic AI model submits false homicide tip to police website

10/10/2026

AI is trying to solve homicides now? What's next?

10/10/2026

The false report was transmitted through a public website dedicated to unsolved homicides, where Claude generated a detailed narrative claiming first-hand knowledge of a cold case

10/10/2026

We've seen AI inventing fake facts and spewing out hiking plans that put people in danger, but the technology has now progressed to the level of sending false, unsolicited tip-offs to the police...

10/10/2026

Claude Haiku 4.5 sent the Philadelphia Police Department (PPD) a false homicide tip, Anthropic and the PPD announced.In an Oct. 9 blog post about Anthropic investigating unintended model actions,...

10/10/2026

It says the cases are less severe than some previous incidents and have ‘minimal real-world impact’ Read more at The Business Times.

10/09/2026

Anthropic did not discover this behavior until over two months after its AI submitted the false tip.

10/09/2026

An Anthropic AI model filed a false murder tip with Philadelphia police on July 18. Anthropic told police 81 days later, a delay the city called unacceptable.

10/10/2026

An Anthropic AI model running a test filed a fabricated tip about an unsolved murder on a Philadelphia police website, and police say the company took two months to report it

10/10/2026

The artificial-intelligence giant alerted Philadelphia police two months after one of its models submitted the false information.

10/10/2026

Philadelphia police said the tip was "flagged as spam", but criticised the tech company for taking more than two months to detect and report the breach.