The Conversation USTech
AI’s false police tip shows agents can go wrong
An artificial intelligence model gave Philadelphia police a false tip, showing how digital agents can follow instructions yet cause harm. An artificial intelligence governance expert warns such errors may become more frequent as these tools take on higher-stakes tasks.
This card is an AI-assisted summary of reporting published by The Conversation US. Read the original report before relying on it for medical, legal, financial, safety, or emergency decisions.