Primary Thought
AI News

OpenAI, Anthropic and Google DeepMind Plan AI Safety Standards Body as Critics Call It a Cartel

The three labs have reportedly approached former Trump AI adviser Sriram Krishnan to lead a FINRA-style self-regulatory body for frontier models, drawing criticism that the proposal could strengthen established AI companies rather than constrain them.

Santanu RC
September 25, 2026
Share
OpenAI, Anthropic and Google DeepMind represented in discussions on industry-led AI safety standards

OpenAI, Anthropic and Google DeepMind are working to establish an industry-run safety body, tentatively called the Frontier AI Standards Agency, that would set testing and reporting standards for advanced AI models without direct government oversight. OpenAI Chief Global Affairs Officer Chris Lehane confirmed during a September 15 Washington briefing that the three companies had been coordinating on safety protocols for several weeks, according to reporting from AI Weekly. Anthropic and Google have not separately confirmed the coordination to reporters.

Modeled on Wall Street's FINRA

The proposed body is modeled on the Financial Industry Regulatory Authority (FINRA), the self-regulatory organization that writes and enforces rules for securities brokers under SEC supervision. According to Forkast News, the structure being discussed would be industry-funded and federally overseen, with frontier models reviewed up to 30 days before release. The proposed approach draws on FINRA's broader model of industry self-regulation under government oversight.

An Unusual Choice to Run It

According to AI Weekly, the three labs have approached Sriram Krishnan to serve as the body's chief executive. Krishnan served as a senior White House policy adviser on artificial intelligence from January 2025 through June 2026. During his time in the Trump administration, he advocated a light-touch approach to AI regulation and opposed the creation of a centralized federal licensing authority for AI. In a July interview with the Financial Times, Krishnan said: “There will not be an FDA for AI.” Krishnan's reported involvement comes after his previous opposition to a centralized government regulator for AI

Proposal Draws Criticism From Rival AI Companies

The plan has drawn criticism from outside the three-lab group. Cohere CEO Aidan Gomez called the proposal “a cartel by any other name,” according to Forkast's reporting. Gomez drew comparisons with the SEC's credit-rating-agency designation process and Europe's Motor Vehicle Block Exemption, arguing that similar frameworks can strengthen established players and create barriers for smaller competitors. Forkast also reported that Meta, xAI, and Nvidia have opposed new government-led AI regulation, highlighting broader disagreement within the industry over how AI safety standards should be governed.

Forkast also questioned FINRA's track record as a model for AI oversight, arguing that its approach to self-regulation allowed markets to expand while reducing political friction and strengthening established players. The publication argues that a similar structure in AI could risk favoring incumbent companies rather than acting as a meaningful constraint on the industry.

A Legislative Alternative Already in Play

The self-regulatory push isn't happening in a vacuum. The FRONTIER Act, a federal bill, proposes a different approach involving independent third-party evaluations of frontier AI developers and their models. Lehane has said OpenAI supports a provision of the bill that would require leading AI companies to allow outside evaluators to assess model safety. That federal proposal would operate alongside OpenAI's support for industry-led safety standards.

When Could the AI Standards Body Launch?

The companies are reportedly targeting a late 2026 or early 2027 launch for the proposed Frontier AI Standards Agency. However, the initiative has not been formally announced, and no public charter or confirmed leadership has been disclosed as of September 2026.

Frequently Asked Questions

What is the Frontier AI Standards Agency?

It's a tentatively named, industry-run body that OpenAI, Anthropic and Google DeepMind are working to establish to set safety testing and reporting standards for frontier AI models, operating outside direct government oversight.

Who would lead the new body?

The three labs have approached Sriram Krishnan, who served as the Trump administration's Senior White House Policy Advisor on Artificial Intelligence from January 2025 to June 2026, to serve as chief executive.

What is the body modeled on?

It's modeled on FINRA, the Financial Industry Regulatory Authority, a self-regulatory organization that writes and enforces rules for securities brokers under SEC supervision.

Why are critics calling it a cartel?

Cohere CEO Aidan Gomez said the proposal amounts to a cartel, arguing that industry self-regulatory bodies like this historically raise barriers to entry and entrench incumbents rather than genuinely constrain the companies that fund them. Meta, xAI, and Nvidia have also opposed the framework.

Is there a government alternative being considered?

Yes. The FRONTIER Act, a federal bill, would create independent, NIST-licensed organizations to verify frontier model safety, rather than relying on an industry-funded and industry-run body.

Related articles

OpenAI AI sandbox escape illustration showing an AI agent bypassing security restrictions, with enhanced network isolation, monitoring and security testing.AI News

OpenAI's AI Sandbox Escape: What Went Wrong and How Security Changed

An OpenAI research agent bypassed sandbox restrictions through a DNS security flaw, prompting an 11-day pause in frontier AI training. Following the September 2026 incident, OpenAI resumed training with stronger network controls, expanded monitoring and additional security testing to prevent future AI agent escapes.

November 3, 2026
Snorkel AI illustration showing $350 million Series E funding, $3.5 billion valuation, AI training datasets, reinforcement learning environments and human expert review.AI News

Snorkel AI's $350M Funding Highlights Demand for AI Training Data

Snorkel AI raised $350 million in a Series E round, reaching a $3.5 billion valuation, nearly three times its previous valuation in 17 months. The company's growth reflects rising demand for specialised AI training datasets and reinforcement-learning environments used to develop advanced AI models.

November 2, 2026
OpenAI Dots always-on AI agents illustration showing cloud computing, connected apps, autonomous workflows and task approval controls.AI News

OpenAI Dots: How Always-On AI Agents Work and Who Can Use Them

OpenAI introduced Dots at DevDay 2026, unveiling always-on AI agents powered by GPT-6 Astra that manage tasks across connected apps. The event also brought the $500-per-month Pro 500 subscription and GPT-6.1 Sol, a lower-cost AI model designed for coding and professional workflows.

November 2, 2026