Major AI Labs Discuss Creating Industry Standards Body

The three leading artificial intelligence companies have been quietly meeting for months to discuss forming an industry standards organization, according to reporting from The Information.

Representatives from Google, OpenAI, and Anthropic have held regular discussions since July about creating a self-regulatory body focused on AI safety testing and auditing. The talks gained momentum following a weekend blog post from Anthropic CEO Dario Amodei calling for AI companies to collaborate on safety measures.

The push for self-regulation

OpenAI CEO Sam Altman told employees at a recent town hall that he supports creating a testing and auditing organization for the AI industry. However, Altman indicated that major AI labs would need to establish such a standards body independently, without waiting for U.S. government backing, according to a source familiar with his comments.

The concept builds on a proposal Amodei published on September 12, in which he advocated for slowing AI development amid safety concerns. His post argued that companies could implement voluntary safety standards while government regulators work on formal AI legislation.

Amodei's position quickly drew support from other technology leaders, including Altman, former Google DeepMind CEO Demis Hassabis, and Elon Musk. Hassabis had previously published his own essay in July proposing a self-regulatory AI body modeled after the Financial Industry Regulatory Authority, which oversees broker-dealers in the securities industry.

Why it matters

The move toward industry self-regulation reflects growing pressure on AI companies to address safety concerns while government frameworks remain in development. A voluntary standards body could establish testing protocols and best practices faster than legislative processes, though critics may question whether companies can effectively police themselves on issues that could limit their competitive advantages. The collaboration among direct competitors also signals that reputational and regulatory risks may be outweighing first-mover advantages in the race to deploy increasingly powerful AI systems.

Adoption depth drives returns

Separate research from PYMNTS Intelligence shows that enterprise AI success depends more on implementation depth than breadth. More than 90% of companies with AI embedded in three or more business functions report seeing returns on their investments. By contrast, only slightly more than half of organizations with AI in just one or two functions see payoffs.

The findings suggest that simply increasing AI spending or deploying tools across more areas does not guarantee results. Instead, deeper integration within specific functions appears to drive stronger financial outcomes.

The Information first reported details of the ongoing discussions among the three AI companies.