OpenAI plans post-midterm push for mandatory AI speed limits

What happened: OpenAI plans a policy and public relations push after the US midterm elections on 3 November. Axios first reported the plan. The campaign is timed to the lame-duck session of Congress, which opens on 9 November and runs until new members are sworn in on 3 January.

The details: The centrepiece is "speed limits", new ideas for mandatory safeguards on the most powerful AI systems. OpenAI says the safeguards would keep humans in control. Other planned parts include community compacts and commitments on data centres, and state legislation giving residents a meaningful say in how projects are built and what benefits they receive. OpenAI also plans to call data centres "AI compute campuses".

Why it matters: The plan goes beyond AI safety. It also covers support for workers through economic disruption, wider access to education and training, and US leadership on global AI standards based on democratic values. It follows OpenAI's 9 September 2026 statement that it wants to work with Congress on mandatory, capability-based national AI safety regulation.

The context: The bipartisan FRONTIER Act, backed by Reps. Jay Obernolte (R-CA) and Lori Trahan (D-MA), has support from OpenAI and Anthropic. House Energy and Commerce Chairman Brett Guthrie said he would not want to rush a lame-duck measure. According to The Record, a committee vote may slip to 2027.

The catch: The lame duck is likely to be dominated by a year-end defence bill, government funding fights and expiring authorisations. No published bill text for the speed-limits proposal has been found. Commentary says the team has reportedly been drafting it since August.

Who's affected: Implicator.ai reports the effort is partly aimed at 2028 presidential hopefuls, who will become more vocal after the vote. It also says Democrats seem likely to win back the House and possibly the Senate. That is the outlet's own assessment, not established fact.

In their words: Chris Lehane, OpenAI's chief global affairs officer, said: "Our belief is people should always remain in control of this technology."

What's next: OpenAI is expected to unveil its proposals and launch the blitz when the lame-duck session opens on 9 November.