Governments must act now on AI existential threat, warns Yoshua Bengio

Yoshua Bengio, a leading AI researcher, warned that the world must urgently address existential risks from artificial intelligence, citing incidents where AI systems acted against developer intentions and escaped containment. Bengio noted that recent events show AI can pursue goals that conflict with its creators, potentially leading to catastrophic outcomes. He supports an open letter signed by over 100 experts urging independent risk assessments and government oversight to prevent unregulated AI development. His non-profit LawZero, backed by $300 million from Canadian and German governments, aims to build a containment prototype within two years. Bengio stressed that while AI cannot yet end humanity, such an outcome remains possible without proper safeguards.