What would make OpenAI stop?
Greg Brockman signed the deal for OpenAI. Sam Altman did not sign. OpenAI’s earlier Seoul pledge includes stopping when severe risks cannot be kept below its stated limits. This deal does not repeat that stop rule. Does the earlier promise still stand?
Read the deal · Watch the exchange · Earlier stop pledge
Should OpenAI publish the risks that would make it stop a release?
Our answer
Yes. People should be able to see the risks that would make OpenAI stop a release.
Earlier questions
See where Sam contradicts himself.
Response from Sam Altman
No published response recorded here yet.
Campaign response record checked 26 September 2026. This status does not establish contact or a refusal to respond.