Anthropic researchers found AI agents can clash, collude, and coordinate in unexpected ways, raising new questions about whether today’s safety tests capture the risks of multi-agent systems.
Related stories
This Is Flock’s AI Search Tool for Cops
Is It Even Possible to Understand Quantum Mechanics?
Smart Toilets Are Already Using AI to Analyze Your Poop
‘Ruin and decay bring a chance of rebirth’: Hidetaka Miyazaki on uniting players in a damaged world
Child sexual abuse survivor alleges Elon Musk’s AI chatbot used photos of her to generate new illegal images
Face Recognition Is Becoming the Norm for Dating Apps