A dramatic shift in the political tide following Tuesday’s midterm elections indicates that the hands-off regulatory approach to AI that AI vendors and enterprises have become accustomed to under President Trump will soon be disrupted, or at least challenged.
The elections clearly showed American citizens expressing their dislike of data centers and fears about AI, voting to restrict or stop data center development in many states and counties.
Moreover, some lawmakers in the newly Democrat-dominated Congress are promising to bring AI leaders to testify under oath, and a bloc of Democratic populists are calling for an AI pause. Abdul El Sayed, the newly elected U.S. senator from Michigan, has called for a “full-on pause” on AI development, while a U.S. representative from Arizona, Yassamin Ansari, is advocating against AI superintelligence. There is also talk of investigations into AI chip export controls, potentially to tighten them after Trump relaxed controls earlier this year.
The results of the midterms reflect consumer sentiment about AI in the U.S. They led to mixed signals for the stock market, with an immediate surge in tech and AI-related stocks the day after the Nov. 3 elections and a sharp decline on Thursday. Declines in shares of AI chipmakers Nvidia, Micron and AMD helped trigger Thursday’s downward turn.
Upset consumers, greater scrutiny
“Voters were mad not just about data centers, but they thought it was the way of stopping AI,” said R “Ray” Wang, founder and analyst at Constellation Research. “They felt AI was going to take away their jobs, and if they could actually do something to stop it, they wanted to go do that.”
Wang argued that misinformation about how AI affects the environment also contributed to the decisive election results, in which voters resoundingly opted for Democrats, in an unprecedented rebuke of Trump and many of his policies.
However, the nationwide ballot results should alert enterprises and vendors that there will be “greater scrutiny of AI companies, particularly around safety, jobs and society,” said Kashyap Kompella, founder and analyst at RPA2AI Research.
While Democrats will likely give greater attention to these issues in hearings, legislative proposals and investigations, federal regulation of AI remains a far way off.
“It is increasingly a wild west of AI regulation, with federal, state and local governments moving in different directions,” Kompella said, adding that an AI application or tool that is acceptable in one state could require additional disclosure, auditing or oversight requirements in another.
How enterprises should respond
Enterprises should expect states to continue to drive regulation, said Michael Bennett, associate vice chancellor for data and AI strategy at the University of Illinois Chicago.
“We’re likely to see continued action there as far as regulation, especially for frontier models,” Bennett said. “Enterprises will need to keep their antenna up.”
Some of the regulations that could affect models could focus on making sure AI models don’t create great risks to consumers and their privacy or affect workers’ jobs, a fear of many. Other regulations could focus on making sure development of frontier models doesn’t lead to higher utility and other bills for consumers.
The right response, then, is for enterprises to treat AI governance as a core business practice, conduct regular AI audits and risk assessments, and have accountability mechanisms in place, Kompella said.
Moreover, as intense consumer vitriol builds around AI data centers, enterprises face greater complexity. Americans, both Democrats and Republicans, voted for more restrictions on AI data center buildouts. And in Texas, home to many planned data centers, the Republican governor has proposed pausing some existing data center approvals.
“What’s really going to happen is we’re going to see at least six months when data center approvals get slowed down,” Wang said, that a slowdown will would hurt the U.S. in the AI race against China and cause a larger pullback of AI use in the country.
“What might have taken less scrutiny for a data center to be built is now going to have so much regulatory red tape,” he said.
“We need AI capacity, and you want to find the balance among national security, local control and what the future looks like,” Wang continued.
Since the animosity against data centers can lead to uncertainty about when infrastructure will be approved and whether enterprises will have the right computing capacity, “enterprises that make significant AI commitments will also need to consider regulatory certainty, energy availability and infrastructure resilience,” Kompella said
For companies that do business with large frontier model providers such as Anthropic, OpenAI and Google, rely on their services and have contractual agreements covering the next few years, it’s important to pay close attention to the financial markets, Bennett said.
“As these investigations almost certainly get underway, it’s hard to imagine that any of the frontier companies are going anywhere, but you certainly could imagine modifications, changes to their policies as a result of investigatory efforts of Congress,” he said. These changes could perhaps slow model releases due to extensive third-party or government review are rolled out.
The market has responded with dramatic volatility to Tuesday’s midterms. Following Trump’s remarks that he would cooperate with Congress, the Dow Jones shot up 1,500 points, its largest single-day gain since April, and the S&P spiked by 1.4 %. The story was different on Thursday, when the three major stock market indexes plunged as investors reassessed signals from the Democratic election sweep.
More competition
Some see an upside, though, to the likely tightening of AI regulation. More of it could lead some states to vie for AI industry investment by offering attractive regulatory environments, energy infrastructure and incentives, Kompella said.
“Competition among states can help establish which regulatory approaches work best,” he said.
He added that another effect is that regulations meant to hold large AI companies accountable could also strengthen them.
“Bigger companies have the resources to absorb compliance costs, navigate complex regulations and relocate infrastructure investments,” Kompella said. “Smaller companies and startups may find it considerably harder. Policymakers will need to strike a balance between protecting the public and preserving competition.”
Esther Shittu is an Informa TechTarget news writer and podcast host and producer covering AI technology.
