OpenAI, Google, and Anthropic in Talks Over Joint AI Safety Standards

Leading artificial intelligence developers OpenAI, Google, and Anthropic have initiated preliminary discussions focused on joint frameworks to tackle rising safety concerns around advanced AI models, an OpenAI spokesperson confirmed to CNBC Business on Tuesday. The talks represent a rare moment of coordination among the technology industry's fiercest rivals, who are under increasing scrutiny from lawmakers and researchers over the potential risks posed by rapid technological advancement.

The foundation for the current dialogues was laid in July when Demis Hassabis, the head of Google DeepMind, published an essay on X proposing a U.S.-led oversight body. Hassabis suggested creating a public-private partnership or a self-regulatory organization backed by federal oversight, modeled after the Financial Industry Regulatory Authority (FINRA), which governs broker-dealers in Wall Street financial markets.

The safety discussions come as competition in the generative artificial intelligence sector reaches unprecedented levels, alongside mounting public and regulatory concern over AI security. On Saturday, Anthropic Chief Executive Officer Dario Amodei issued a high-profile plea encouraging major AI developers to deliberately slow the pace at which they push the capabilities of their most advanced models forward. Amodei’s call received immediate public backing from executive leaders across rival firms, including Hassabis and OpenAI Chief Executive Officer Sam Altman.

Praising the essay written by Anthropic's leader, Hassabis stated in a post on X that "Dario's essay points towards the right path forward." The DeepMind chief added that while "the details need working through, but the direction is correct for meeting this critical moment." Altman echoed those sentiments in a separate post on X, confirming that pacing the release and capability growth of frontier models has become a "primary topic" inside OpenAI in recent weeks, promising the company would share more details "soon."

The effort to align on voluntary standards aligns with recent statements from OpenAI’s policy leadership. Earlier this month, Chris Lehane, the company’s chief global affairs officer, authored a blog post detailing OpenAI's intention to advance "industry-led standards." Lehane noted that OpenAI is actively working to establish a voluntary oversight framework alongside peer technology companies, emphasizing that such efforts should proceed "with or without government support."

Lehane clarified in his publication that voluntary industry commitments are not intended to substitute for formal government intervention or legislative mandates. "Any industry-led standards would complement — not replace — mandatory federal safeguards and democratic oversight," Lehane wrote. However, he stressed the urgency of the moment, writing that "we have entered a different phase of AI capabilities, and everything should be on the table — including considering what can be built voluntarily outside government as an immediate step."

The discussions among OpenAI, Google, and Anthropic were initially reported by technology news site The Information before being confirmed by an OpenAI representative to CNBC Business. Requests for comment sent to Google and Anthropic regarding the ongoing safety deliberations were not immediately returned on Tuesday.