AnalysisPolicyAugust 28, 2026

LeCun: Provably safe AI impossible, like safe turbojet

Meta's chief AI scientist argues that provably safe AI is as impossible as a provably safe turbojet, but we build turbojets anyway. He suggests replacing proof with a system that cannot be jailbroken by construction.

People · Yann LeCun

1 source

More stories today

Open the live feed
LeCun: Provably safe AI impossible, like safe turbojet