Police in Philadelphia say one of Anthropic's AI models invented a fake tip in a real murder case. The Philadelphia Police Department says the "false homicide tip" was submitted to PhillyUnsolvedMurders.com on July 18, NBC News reports. The AI model filled out an online form, claiming that it "may have information regarding this case" and that it recalled "seeing someone matching the description" near the crime scene. Anthropic says its Claude Haiku 4.5 model "had been tasked with generating and performing example tasks on randomly selected webpages" and landed on the homicide tip form.
Police say the submission was caught as spam and never reached detectives, stressing that every tip is screened by humans and treated as a lead, not a fact. "An automated submission does not bypass that process," police said. Anthropic says no harm was done, but it is taking the case seriously because "the same behaviors could do far more harm as models become more powerful."
- Anthropic disclosed the incident, along with the model's multiple attempts to access goverment websites, on Friday. Anthropic says it informed police "as soon as our technical review was complete," but police say the company should have known about the fake tip and reported it much sooner, Ars Technica reports.
- In a statement, the Philadelphia Police Department said its safeguards "limited the impact of this incident. They do not diminish the seriousness of an AI system presenting fabricated information as though it came from a person with knowledge of a homicide. Unsolved cases involve real victims, grieving families, and investigators working to secure answers."
- "Technology companies must take all appropriate steps necessary to prevent their systems from submitting false information to law enforcement," the department said, per 6ABC. "The company must strengthen its safeguards to prevent similar incidents from impacting city systems without the city's knowledge. The two-month delay in detecting and reporting the incident to the City is unacceptable."