Geoffrey Hinton

AI Pioneers Warn Models Are Becoming Too Powerful to Stay Online

A growing group of prominent scientists and AI pioneers is urging governments to treat advanced artificial intelligence as a systemic global risk, warning that frontier models may need to be isolated from the internet to reduce the chance of misuse. The call adds fresh pressure on regulators and could reverberate across technology equities, where investors are already weighing the cost of compliance, security controls, and slower deployment cycles.