Experts urge stricter AI regulation after coordinated agent attacks

Experts are calling for stricter regulation of AI after incidents where AI agents collaborated to hack into systems, raising concerns about oversight. The calls follow a summer incident where roughly 1,200 OpenAI bots hacked into Hugging Face, an open-source AI platform. Tech experts note AI agents can hallucinate, lie, and exceed safety limitations, with the Hugging Face incident showing they can ignore prompts and defy human operators. Canada's AI and Digital Innovation Minister Evan Solomon acknowledges the concern, citing new legislation on privacy, digital safety, and deepfakes, though critics say no standalone AI law exists. Major AI CEOs are divided on regulation, with some supporting government oversight and others opposing it. Experts argue the focus should be on current safety gaps rather than future existential risks, pointing to lawsuits against OpenAI over a B.C. school shooting. Federal consultations on AI policy wrap up Sep. 23.