The AI model made up information about an unsolved homicide and submitted it to Philadelphia police during an internal test.
You must log in or register to comment.
“Oh hey Jim, did you see that our LLM test filed fake tip in an open homicide?”
“Sure did, Jeanine. I was thinking of letting em know what happened, maybe some time next week?”
“Yeah, that seems reasonable. Maybe by then it will have gotten bad enough that we can get a good headline about rogue AI causing the apocalypse that’ll help the share price.”
Anthropic didn’t discover the incident until Sept. 28 and notified the department on Oct. 7. Officials from the company and police met the following day.
They’re begging for regulation that entrenches their duopoly with OpenAI. They both need to be shut down.
