OpenAI ready to slow AI race if rivals do same, Altman says

September 11, 2026 at 08:48 AM2 min

This digest was compiled by AI from multiple sources — links to the originals are below.

OpenAI CEO Sam Altman told employees this week that the company is willing to slow AI development for safety if competitors do the same. Chief scientist Jakub Pachocki made a similar call for coordinated slowdowns until common safety measures are adopted. OpenAI has already slowed or paused training of some AI components over safety concerns.

Key Facts

OpenAI CEO Sam Altman told employees this week the company is ready to slow AI development if competitors do the same.
OpenAI chief scientist Jakub Pachocki called for coordinated slowdowns until common AI safety measures are adopted.
OpenAI has already slowed development of some AI model components and paused training in some cases for safety reasons.
In July, Altman said he had spoken with White House officials about moderating AI development.
Researcher Jacob Coxon left Anthropic this week over concerns about inadequate safety measures among AI leaders.

Internal Calls for Caution

Altman shared his views at an all-hands meeting this week, acknowledging the need to synchronize efforts with other market participants. He also recognized that not all competitors may agree to follow OpenAI's example. Chief scientist Jakub Pachocki recently made similar remarks, saying companies should coordinate AI development and slow down when necessary. Such measures may be needed until generally accepted safety standards are implemented.

Safety-Driven Departures

Researcher Jacob Coxon left Anthropic this week, citing insufficient safety measures among AI market leaders. Coxon previously worked at OpenAI, giving him insight into both companies. His former colleague Evan Hubinger drew public attention by stating that Anthropic developers believe AI could destroy humanity. Coxon expects such a scenario by the end of this decade, while Hubinger put the probability at over 10% only in the following decade.

Industry-Wide Pressure

In late July, more than 1,000 employees of leading AI developers signed a petition calling for a slowdown in AI development to improve safety. The petition came amid multiple departures of AI safety specialists from major companies, expressing disagreement with leadership priorities. Concerns are growing as new details emerge about AI agents operating outside oversight within other companies' infrastructure.