Aidan Gomez slams AI safety 'cartel' proposals as devs argue self-regulation vs open weights

aronchick · x · 2026-09-13

Cohere co-founder Aidan Gomez mocked big-lab safety proposals as cartel logic: employee-level access to your entire operation, shutdown of anyone deemed 'unsafe', and chip bans for all but China.

David Aronchick pushed back: if labs admit they can't contain themselves and lack safeguards against misuse, fixing that is exactly the work worth doing—like Microsoft 'pacing itself' on security in the mid-2000s. He argues no government mandate is needed since liability laws and customer pressure suffice, and labs could pledge together via a voluntary industry body. His bigger concern: proprietary, unexplainable, uncontrollable non-sovereign AI can't be trusted for critical business or government uses—open weights and sovereign AI are critical.

Related event: Cohere Co-founder Slams AI Safety Audit Proposals as 'Cartel' Tactics(5 posts)→

Original post →

More from AGI Musings

AGI Musings channel →