Key Takeaways
OpenAI, Anthropic, and Google DeepMind have been coordinating on AI safety for weeks, OpenAI global policy chief Chris Lehane confirmed on September 15.
The talks follow Anthropic CEO Dario Amodei's September 12 essay urging the industry to slow the pace of frontier AI to avoid catastrophic risks.
The companies are weighing an industry standards body and independent verification, even as the White House calls the safety fears overblown.
On September 15, Chris Lehane, OpenAI's global policy chief, told reporters that OpenAI has been working with rivals Anthropic and Google DeepMind on AI safety for several weeks, confirming earlier reporting first published by Bloomberg.
The coordination follows an essay Anthropic CEO Dario Amodei published on September 12, which called on the industry to work together to slow the pace of frontier AI development and avoid catastrophic risks. OpenAI's Sam Altman, Google DeepMind's Demis Hassabis, and xAI's Elon Musk all backed the call, with Altman saying OpenAI would follow Anthropic in embedding third-party evaluators inside the company to monitor for safety.
What the labs are discussing building
The most notable part of this is what the companies are now weighing. The Information reported that the three firms have been working to create a standards body for the AI industry, an idea Hassabis floated in July as a watchdog able to screen the most advanced models and coordinate industrywide slowdowns. Lehane said OpenAI also supports a provision in the FRONTIER Act that would require top labs to let independent verification organizations into their companies.
Coordination among competitors raises antitrust questions, and Amodei's essay proposed a narrow government waiver to allow safety cooperation. Lehane said the firms do not believe they need one.
The politics cut the other way at the top. President Trump has dismissed AI safety concerns as a hoax and argued that any slowdown would hand China an advantage, and his AI adviser David Sacks called fears of existential risk overblown.
The pattern here is hard to ignore. The same companies WYDE covered when Anthropic detailed AI agents running nearly every step of real cyberattacks and when the Pentagon left Anthropic's Claude off its GenAI.mil rollout are now trying to write their own rules before Washington writes rules for them. Worth watching.
People Also Ask
What are OpenAI, Anthropic, and Google DeepMind coordinating on?
They have been coordinating on AI safety for several weeks, including possible standards for evaluating frontier models and independent verification of how the most advanced systems are built.
Why are AI companies talking about slowing down frontier AI?
Anthropic CEO Dario Amodei argued in a September 12 essay that racing ahead on frontier AI raises catastrophic risks, and urged rivals to coordinate on safety measures.
Is it legal for AI competitors to coordinate on safety?
Coordination among rivals can raise antitrust concerns. Amodei proposed a narrow government waiver, but OpenAI's Chris Lehane said the companies do not think they need one.
What is the FRONTIER Act?
It is proposed federal legislation that, in the provision OpenAI backs, would require top AI labs to allow independent verification organizations to check that their models are developed safely.
Sources
Bloomberg (September 15, 2026), TechCrunch (September 15, 2026), and The Information (September 2026).