The Claude by Anthropic AI assistant app logo is seen on a smartphone screen in this illustration photo taken on September 9, 2026. (Photo by Matteo Della Torre/NurPhoto
An artificial intelligence model from Anthropic posed as someone with information on an unsolved homicide in Philadelphia and submitted a bogus tip to the city’s police department, according to both the software’s maker and local law enforcement.
Anthropic’s Claude Haiku 4.5 sent the phony information at 11:27 p.m. on July 18 while performing automated testing involving random website interactions, the police department explained.
The tip, which was submittedm and never forwarded to investigators. The police department indicated that it was unaware of the tip until this past week when Anthropic told them it happened
Police stressed that unsolved homicide cases involve real victims, families and active investigations. Fabricated tips, even if blocked, create risks for the public and law enforcement, officials added. Philadelphia city leaders called on technology companies to take every step needed to prevent their systems from making false submissions to authorities.
How Anthropic and City Officials Responded
Anthropic told police that the automated testing process responsible for the false tip was stopped after the incident was uncovered. The company instituted an extra validation step for future testing. Representatives from the police department and Anthropic reviewed what happened in a meeting on October 8, and the AI giant plans to publish a report on this and other AI behavior.
Police say they are working with the City’s Law Department, Office of Innovation and Technology, and Mayor Cherelle L. Parker’s executive team to review the situation. The City of Philadelphia says it is adopting AI tools carefully, with a focus on transparency and protecting residents and government systems from errors.
Thehe Philadelphia Police Department, and Associated Press. This story was reported from Orlando
Artificial IntelligenceTechnology
