Three questions that will define AI's political power

AI becoming a political issue means candidates and groups are asking voters to judge campaigns partly by their positions on artificial intelligence. The 2026 midterms are the first major test of whether AI policy can influence elections. The results may shape strategies for the 2028 presidential campaign cycle.

Some groups are making regulation itself the campaign message. Guardrails Alliance launched ads attacking Republicans and Democrats over their AI positions in four states. Abdul El-Sayed called AI the most important issue before Election Day and linked it to jobs and affordability. Democratic group Tomorrow Together also attacked Republicans’ AI record in several states.

The test is whether these messages win votes or create political risk. Other AI groups spent heavily but mostly avoided mentioning AI in their ads. Results will show whether explicit AI campaigning is viable and whether candidates can afford vague or incomplete agendas.

Several types of political actors are making AI policy part of their message. Pro-regulation groups want candidates judged on issues such as safeguards, transparency, and limits on powerful systems. Candidates aligned with that approach are presenting AI as connected to jobs, affordability, and government oversight.

Guardrails Alliance launched ads targeting Republicans and Democrats in Arizona, Montana, North Carolina, and California. Tomorrow Together attacked Republicans’ AI track record in Michigan, Wisconsin, Arizona, and Pennsylvania. Abdul El-Sayed made AI central to his Michigan Senate campaign. Other candidates, including Manny Rutinel and Scott Wiener, are associated with major AI legislation and transparency efforts.

The actors do not all use the same message. Public First and Leading the Future spent heavily through affiliated super PACs, but their ads largely avoided AI. Musk-backed campaigns provide another model, using his money and political connections as either support or an attack point.

The article identifies two major streams of AI-related political spending. Public First and Leading the Future spent more than $50 million through affiliated super PACs in nearly 50 races. Their spending sought political influence connected to AI, even though their advertisements usually did not mention the technology.

Elon Musk’s political spending was much larger. He contributed more than $400 million and was the biggest donor in the midterms by far. His involvement also included DOGE and ties to the Trump administration. Musk-affiliated spending appeared in contests including the Ohio Senate race and the Texas Senate race.

The amounts show that AI politics is not limited to policy advertisements. Money can support candidates, influence races, or create political targets. The election results may indicate whether Musk’s spending elects allies and whether large AI-linked expenditures translate into visible political power.

The groups are using different political strategies. Guardrails Alliance and Tomorrow Together are treating AI regulation as a voter-facing issue. Their ads criticize candidates’ records and positions, aiming to make AI part of campaign debate. This approach tests whether voters will respond to explicit policy arguments.

Public First and Leading the Future chose a quieter route. They spent more than $50 million through affiliated super PACs in nearly 50 races, but their television and digital ads largely avoided mentioning AI. The key mechanism is indirect influence: spending can affect candidate contests without asking voters to focus on AI policy.

The contrast matters because the midterms will compare these approaches. Direct campaigning could show that regulation motivates voters. Indirect spending could show that groups prefer influencing who wins without making AI a prominent public issue. The results will help shape future strategies.

The 2026 results will give campaigns evidence about two untested strategies. One is explicitly campaigning for or against AI regulation. The other is aligning with Elon Musk or using his political involvement against opponents. Both strategies could influence candidate selection, advertising, fundraising, and issue priorities before 2028.

For example, Abdul El-Sayed used Musk’s political involvement to attack Republican Mike Rogers. In Ohio, Republican Jon Husted benefited from spending by Musk’s America PAC. In Texas, Musk-affiliated spending became part of the contest involving James Talarico and Ken Paxton. These races connect policy messaging, billionaire spending, and candidate performance.

Success could encourage more candidates to make AI central to their platforms or seek Musk-linked support. Failure could make explicit regulation messages or Musk alignment seem risky. Competing AI factions will claim victories, explain losses, and draw lessons about what worked ahead of 2028.

Congress could decide whether stronger AI safeguards gain a louder national platform. The article describes candidates who support a Cabinet-level AI agency, a ban on super intelligence, transparency legislation, limits on data center construction, and bans on nondisclosure agreements. Their election would help determine the size of a congressional bloc pushing back against efforts to limit regulation.

The candidates represent different policy tools. Abdul El-Sayed wants a new federal agency and a ban on super intelligence. Manny Rutinel sponsored the nation’s first comprehensive AI bill. Scott Wiener’s transparency legislation was central to an industry-dividing regulatory fight. Matt Little supports national limits on data centers and nondisclosure agreements.

Their victories could produce a more vocal AI safety coalition in Congress. That coalition could shape regulatory battles in 2027 and beyond. Its size and influence would indicate how far stronger safeguards can advance nationally.

AI regulation is the creation and enforcement of public rules for developing and using artificial intelligence. Such rules can address safety, transparency, powerful systems, data centers, government agencies, and agreements that restrict what workers can disclose. The article presents these as active political questions rather than purely technical decisions.

The need for government action follows from the scope of the choices. Abdul El-Sayed wants a Cabinet-level agency and a ban on super intelligence. Scott Wiener backed transparency legislation. Matt Little supports national bans on data center construction and nondisclosure agreements. These proposals use public authority to set common requirements or limits.

Technology companies can develop systems and adopt voluntary practices, but they do not alone decide national laws or public spending rules. Government decisions can apply across industries and jurisdictions. The article shows that candidates and Congress may determine whether stronger safeguards advance in 2027 and beyond.