Anthropic has revealed that one of its AI models submitted a false tip to the Philadelphia Police Department claiming knowledge of an unsolved homicide. The AI startup didn’t inform police of the incident for several months.
\n\nThe Hill reports that Anthropic’s Claude Haiku 4.5 filled out a form on Philadelphia Police Department website used to submit tips on unsolved crimes to detectives. The AI claimed to have personal knowledge of who committed a murder, a case of AI hallucination.
\n\nAnthropic said it had set the model loose on “example tasks on randomly selected webpages.” The rules blocked a handful of actions: no creating accounts, no submitting personal information, no making purchases. Nobody told it not to submit forms to the police.
\n\nThe false tip got flagged as spam and investigators never followed up on it. But Philadelphia police said Anthropic didn’t tell them what happened until last week, months after the July tip.
\n\nThe department called that delay “unacceptable.” Its statement didn’t stop there: “Unsolved cases involve real victims, grieving families and investigators working to secure answers,” it said, adding that “technology companies must take all appropriate steps necessary to prevent their systems from submitting false information to law enforcement.”
Original source: https://www.breitbart.com/tech/
