Legal guardrails and company best practices for AI safety should require that powerful AI developers demonstrate safety to regulators before building systems approaching AGI, and should not proceed if safety cannot be demonstrated.

normativepending

Speaker

Yoshua Bengio

Evidence Quote

“if they can't demonstrate to the public to the regulator that their system is safe enough then they shouldn't even build it”

Source

Why a Forefather of AI Fears the Future | World Science Festival— World Science Festival
Created: 8/12/2026, 6:45:56 PM

Your notes

Loading notes...