The US midterm elections on November 3 are unlikely to produce a sweeping federal AI law, but they could shape several fights that matter to the AI industry. CNBC reports that the next Congress could set the direction of the AI buildout, with about $5 trillion in capital spending through 2030 at stake as public opposition to data centers grows.
Where the race stands. As of October 21, Decision Desk HQ gave Democrats a 77% chance of winning the House and 55% in the Senate, while Kalshi traders put those odds at 91% and 65%. Not every model agrees: one forecast still gave Republicans a 64% chance of keeping the Senate. Republicans currently hold the Senate 53-47. The Senate is the tighter fight, and forecasts can move before election day.
1. Whether states can keep writing AI laws. In July 2025 the Senate voted 99-1 to strip a ten-year ban on enforcing state AI laws from a budget bill, and Trump later issued an executive order threatening some federal broadband funds for states with AI laws his administration considered onerous. States had enacted 109 AI laws and 28 data center laws as of July 1. Experts say a standoff over federal preemption awaits the next Congress whichever party wins, and that Trump is unlikely to sign a bill that leaves state laws standing.
2. How much money the industry spends to shape the answer. AI-related political spending has topped $300 million this year, with three PAC networks that favor different regulatory approaches and have spent against each other in at least six primaries. Leading the Future, backed by a16z’s founders and OpenAI’s Greg Brockman, wants one national framework in place of state rules, and Anthropic’s contribution is restricted to policy education. The industry isn’t a single bloc.
3. Data centers. Data centers have become a bipartisan issue: New York’s governor imposed a one-year statewide moratorium on new hyperscale data centers, Texas’s governor froze the first stage of approvals for hundreds of projects, and the Trump administration got seven major tech companies to pledge to pay for the power and grid costs of their data centers. JPMorgan says one Pennsylvania House race alone has consequences for more than a dozen proposed data centers. Sanders and Rep. Ocasio-Cortez back a federal moratorium on data center construction until national safeguards are in place.
4. Proposals like a national AI wealth fund. Sanders’ American AI Sovereign Wealth Fund Act would impose a one-time 50% tax, paid in stock, on the biggest AI companies. An independent commission would use the voting shares to block harmful decisions, and Sanders puts the fund’s value at about $7 trillion at current valuations. Critics, including George Mason law professor Ilya Somin, argue it is almost certainly unconstitutional. Others say it would disadvantage US companies and push them offshore. Supporters note that OpenAI and Anthropic have themselves floated wealth-fund ideas. Forbes calls it a long shot, contingent on Democrats winning control. Even then, a bill needs the president’s signature or a two-thirds vote to override a veto.
The bipartisan track. Sens. Cruz, Klobuchar and Thune confirmed work on a bill covering catastrophic AI risks with voluntary standards and narrow preemption of stricter state laws, and prediction markets gave 63% odds that some AI legislation passes this Congress. That’s a market estimate, not a prediction of what the bill contains.
Bottom line: The midterms are more likely to shape the AI industry’s operating environment than to rewrite its rules. Who controls committees will influence which bills get hearings, the preemption fight carries into the next Congress either way, and data center politics are already moving at the state level. What isn’t settled is whether any of the larger proposals, such as the wealth fund, can become law.
