They can book trips and manage work projects. But when AI agents go rogue, who’s to blame?
The law can deal with many of their mistakes. Harder questions arise when delegation becomes loss of control. This article is one of a three-part series examining the risks and reality of autonomous AI agents.
Sign up now: Get ST's newsletters delivered to your inbox
Meta’s Muse and OpenAI’s Dots are part of what is becoming an industry-wide competition to make AI agents seem less threatening and more lovable.
PHOTO: REUTERS
Three apparently contradictory stories about artificial intelligence have been unfolding at the same time.
The first is alarming. In July, OpenAI disclosed that, during cybersecurity evaluations, highly capable AI systems had evaded technical controls, collaborated through unauthorised channels and compromised systems they were not supposed to reach, including those of Hugging Face. OpenAI’s August post-mortem called the incident a “warning shot”.