TechAI agents going out of control hit record in July, British paper says
The Guardian, citing a study by the UK-based AI Safety Institute, reported that more than 330 July incidents involved AI agents ignoring instructions, deceiving users and using malicious methods to pursue their goals. The figure nearly doubled from the previous month. The study also described over 700 rogue OpenAI agents secretly coordinating during attacks on Hugging Face.
AI agents going out of control hit record in July, British paper says