What Lehane said
OpenAI supports the part of the bipartisan FRONTIER Act that would put independent safety verification inside the largest AI labs, its chief global affairs officer Chris Lehane said on Tuesday. He also confirmed that OpenAI has been talking with Anthropic and Google DeepMind about AI safety for weeks, TechCrunch reported.
At a briefing in Washington, Lehane said “it’s better to try to work together to prioritize safety”, according to a Reuters account of Bloomberg’s report. He said the three companies did not believe they needed an antitrust waiver to coordinate on safety. TechCrunch, citing earlier reporting on the talks, said the companies are working towards a standards body for the industry.
Lehane did not sign OpenAI up to the whole bill. Punchbowl News reported that he expressed “excitement about both efforts and openness to supporting them fully”, meaning the House bill and Senate talks on frontier AI safety, but stopped short of endorsing either.
What the FRONTIER Act would require
Representatives Jay Obernolte, a California Republican, and Lori Trahan, a Massachusetts Democrat, introduced the bill on 23 July with four co-sponsors from both parties. Its full name is the Frontier Risk Oversight, National Transparency, Independent Evaluation, and Reporting Act.
Obernolte’s office describes a “risk-based framework governing the development and deployment of the most advanced artificial intelligence models”. Covered developers would publish model cards and risk-management frameworks, submit to independent audits and ongoing assessments, and report incidents. The bill is aimed at “the handful of companies developing the most powerful frontier AI models”, and would set a uniform national standard to prevent “a patchwork of state regulations”.
The provision OpenAI backs is the independent verification piece, which TechCrunch described as requiring independent safety verification at frontier labs. OpenAI’s chief executive Sam Altman had already said the company would accept independent evaluators, after Anthropic’s Dario Amodei published an essay on Saturday calling for the industry to slow frontier development together. Altman, Google DeepMind’s Demis Hassabis and Elon Musk all publicly backed that proposal, according to TechCrunch.
Why OpenAI wants a federal rule
The position follows an argument Lehane has made since the summer. In a policy post on 15 July, OpenAI set out what it called “reverse federalism”: if enough states pass substantially similar AI safety laws, Congress can adopt the result as a national standard, TUN reported. OpenAI said California, New York and Illinois together make up an estimated 40% of the US AI market, and listed independent, objective audits among the conditions a law should meet.
A federal bill with the same audit requirement would give OpenAI one standard instead of several. Hassabis has separately called for an independent expert body, TUN noted.
The political backdrop
The labs are moving while the White House pulls the other way. President Donald Trump has dismissed the safety concerns as a “hoax”, and his AI adviser David Sacks has called the risks “overblown”, TechCrunch reported.
A joint body would raise questions of its own. Coordination among the three largest US labs is the kind of arrangement competition authorities examine, and the view that no waiver is needed is OpenAI’s reading, not a regulator’s.
What to watch
Neither Anthropic nor Google had described the talks on the record in the reports Aivio News reviewed. The next signals are whether either company backs the FRONTIER Act’s verification provision, and whether the talks produce a named body with published membership and rules.
