The midterms are over. The bills are not written yet, but the list is drawn. Democrats are preparing a run of legislation on artificial intelligence and data centres for January, when they take control of Congress, according to lawmakers, congressional aides and advocates who spoke to Reuters on Friday.
Three strands are visible. First, shielding consumers from electricity costs tied to data-centre construction. Second, requiring independent testing of advanced AI systems against safety standards. Third, clarifying when AI companies can be held liable for harm their products cause. Greg Casar of Texas, chair of the Congressional Progressive Caucus, said this week that AI executives should already be testifying under oath.
For Republicans this is a break in continuity. Their Congress largely focused on promoting AI and declined to pass legislation restricting it, in a year when voters turned against rapid data-centre development and its power demands. The shift in tone is not only partisan; it follows the mood the ballot box recorded.
The odds remain long, for a structural reason: Donald Trump is still in the White House and has repeatedly dismissed calls for mandatory AI rules. So Democrats are leaning less on clean packages than on committee control — hearings, subpoenas, investigations, and vehicles that survive a veto. A veto can be overridden with two-thirds majorities in both chambers; alternatively, AI rules can be attached to the annual must-pass bills. "The probability of legislation is much higher than people anticipate," said Ed Mills, who manages policy research at Raymond James.
Names and jurisdictions are already assigned. Chuck Schumer, expected to become majority leader, named protecting consumers from data-centre-linked energy costs as a priority. Lori Trahan of Massachusetts plans legislation requiring independent experts to evaluate leading models against safety standards, and a bill clarifying corporate responsibility when AI systems act harmfully, her deputy chief of staff said. Zoe Lofgren, seeking to chair the House Science Committee, promises "responsible regulations for AI while also fostering innovation." Abdul El-Sayed, who won a Michigan Senate seat on Tuesday, campaigned on an independent AI regulator with powers to police developers.
Pressure on the industry is already concrete. In October, Lofgren and other Democratic committee leaders asked OpenAI, Anthropic, Meta, Google and SpaceX to detail their security practices and report instances of their AI tools taking actions their designers did not intend.
For markets this is no longer an abstraction. Technology shares sold off sharply on Thursday, on concerns that a new Congress will slow AI development and on doubts about whether the capital expense of the buildout will pay off. A crisis, one policy analyst suggested, could build enough support to overcome presidential opposition. Until then the inventory is proposals, hearings, and the question of how much of it becomes law in January.
