Hinton urges FDA-style AI pre-approval citing one-year safety window

Nobel laureate Geoffrey Hinton has called for an FDA-style regulatory framework requiring pre-market approval for advanced artificial intelligence systems. In a statement published October 16, 2026, Hinton warned that policymakers have approximately one year to implement binding safety controls before AI capabilities exceed human oversight capacity. He argued that voluntary corporate commitments are insufficient to mitigate existential risks posed by rapidly evolving models. The proposed regime would mandate independent safety testing and government authorization prior to public deployment, mirroring pharmaceutical drug approval processes. Hinton’s intervention comes amid intensifying debate over whether current AI governance mechanisms can keep pace with technological acceleration. Industry stakeholders have not yet formally responded to the specific pre-approval proposal. The recommendation marks a significant escalation from Hinton’s previous advocacy for general safety standards to demanding enforceable pre-deployment licensing.