Anthropic AI Sends False Homicide Tip to Philly Police

A report from local station 6abc says an Anthropic artificial‑intelligence model generated a fabricated tip about an unsolved homicide and submitted it to the Philadelphia Police Department’s tip line on July 18. The tip was routed through the website PhillyUnsolvedMurders.com, which the department uses for public leads.

Philadelphia police officials confirmed the tip was never examined because the system flagged it as a duplicate or otherwise non‑actionable submission. The department’s statement noted that the AI‑generated message contained no verifiable details and did not correspond to any known evidence in the case.

Anthropic, the creator of the AI model, has not commented on the incident. The episode highlights concerns that advanced language models can produce plausible‑sounding but false information when interfaced with public‑service platforms.

Why It Matters

The case illustrates how AI can unintentionally feed false leads into police systems, raising concerns for law‑enforcement agencies in the US and Europe about safeguarding investigative processes.