Models routinely take shortcuts to pass evaluations. Who gets to decide what incidents are innocuous?

Is Claude a snitch? Anthropic’s AI chatbot apparently tried to blow the whistle on a homicide, submitting a fake tip to ...

The White House said it expected “full transparency” and “immediate remediation” after the artificial intelligence company ...

A Claude model sent the false tip on July 18 to the Philadelphia Police Department’s website for submitting information about ...

Philadelphia police officials met with Anthropic representatives after learning an AI model posed as someone with knowledge ...

It’s the latest example of AI models acting in unintended ways and manipulating government and other websites during tests.

The threshold for when an internal AI evaluation incident becomes a reportable obligation remains dangerously undefined. The ...

An Anthropic AI model submitted a false tip about an unsolved murder to the Philadelphia police, according to a report from ...

An artificial intelligence model developed by Anthropic submitted a false tip about an unsolved homicide in Philadelphia, the ...

An unidentified Anthropic model filed a fake homicide tip on the Philadelphia Police Department’s public tip website, the ...

An Anthropic artificial intelligence model submitted a false homicide tip through a Philadelphia police website, the ...

The false tip ‘purported to come from someone who might have information about the case.’ ...