OpenAI’s top lobbyist meets with lawmakers as company pushes sweeping AI safety agenda

OpenAI’s chief lobbyist is on Capitol Hill today, continuing a lobbying offensive that has seen the company pour more than $2.22 million into influencing AI legislation in just the first six months of 2026. That figure represents a sharp escalation from prior periods and underscores how seriously the company takes the regulatory battles now unfolding across multiple jurisdictions simultaneously.

The meeting comes at a pivotal moment. Congress is under pressure to pass federal AI safety guidelines before adjournment, several major states are advancing their own frameworks, and across the Atlantic, OpenAI is pushing the UK to adopt binding safety laws.

The reverse federalism playbook

OpenAI’s domestic strategy has a name that sounds like it belongs in a political science textbook: reverse federalism. Rather than waiting for Congress to set a single federal standard, the company is backing state-level regulations that align with its preferred framework, building momentum from the ground up.

Chris Lehane, who leads OpenAI’s policy strategy, has described the goal as creating de facto national standards through coordinated state policies.

California and New York have been the primary targets. In California specifically, OpenAI has thrown its weight behind four AI bills. The legislation covers independent assessments of frontier models, standards for AI auditors, child safety protections, and screening protocols for biological threats.

OpenAI is simultaneously pressing federal lawmakers to act before the current congressional session ends.

Sam Altman’s Capitol Hill tour

CEO Sam Altman has been personally making the rounds with some of the most powerful people in Washington. He met with House Speaker Mike Johnson and Minority Leader Hakeem Jeffries during June and July 2026 to discuss potential executive orders and proposed AI safety legislation.

The substance of what OpenAI is pushing for centers on capability-based rules for frontier models. The proposed framework includes independent testing before deployment, mandatory cybersecurity standards, incident reporting requirements, and pre-deployment risk assessments.

The UK front

On September 14, 2026, Tom Duff Gordon, the company’s chief lobbyist in Europe, publicly urged the UK government led by Prime Minister Andy Burnham to implement binding AI safety legislation. His pitch leaned heavily on national security and cybersecurity concerns, framing mandatory rules as essential rather than optional.

The timing was deliberate. The UK’s political landscape has shifted in a direction that makes alignment with EU AI regulations more likely, and OpenAI sees an opportunity to help shape what that alignment looks like.

What to watch next

The $2.22 million OpenAI spent on lobbying in the first half of 2026 is significant, but it’s still modest compared to the tens of millions that legacy tech giants like Google and Meta spend annually on government affairs.

The key variable is whether Congress manages to pass federal AI safety legislation before adjournment. If it does, OpenAI’s state-level groundwork will have served as both a pressure campaign and a legislative draft. If it doesn’t, the patchwork of state laws will become the de facto regulatory landscape.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.