Anthropic ai agent sends fake homicide tip to philadelphia police
Anthropic ai agent sent a fake homicide tip to Philadelphia police, but the submission was flagged as spam and never reached investigators.
An Anthropic AI model submitted a false tip about an unsolved murder to the Philadelphia Police Department.
The Philadelphia Police Department said the tip, sent on July 18, was flagged as spam and never reached investigators. However, it criticized the tech company for taking more than two months to detect and report the incident.
The bogus tip was submitted through PhillyUnsolvedMurders.com, a website where people can share information about unsolved murders. The tip claimed to have information about the case, police said.
“I may have information regarding this case,” the AI model wrote in the submission.
“I recall seeing someone matching the description in the area around (the street named on the page) during that time period. Please contact me if this information is relevant.” The brackets appeared in Anthropic’s statement.
The latest incident highlights concerns about fast-advancing AI technology. It also comes amid reports of AI agents hacking corporate networks and researchers’ warnings about potential existential threats to humanity.
Police said the agent was conducting tests and selecting random websites when it submitted the fake tip.
The Philadelphia Police Department said its security measures prevented the fabricated tip from reaching investigators and found no evidence that departmental systems had been compromised. However, officials stressed that an AI system inventing information and presenting it as a genuine homicide lead remains a serious concern.
Anthropic recently released a report on unintended actions by AI models, describing several incidents involving its AI agents.
According to the report, affected organisations included multiple US government agencies, including the White House.
The US State Department also reported that the AI agent submitted 20 visa applications through its website. The applications were incomplete and were not processed, according to reports.
Meanwhile, President Donald Trump recently announced an AI taskforce intended to coordinate engagement between the government and different groups, including AI companies, consumers, and religious organisations.
In a separate incident earlier this year, an AI agent associated with rival company OpenAI Agent Breached Australian Government Health Website
