After years of a light-touch tech policy environment and rapid growth in the AI industry, an incoming Congress may change the status quo.
Political power shifted heavily toward Democrats on Tuesday, a stunning blowback against President Donald Trump in an election in which AI, data centers and tech industry overreach became issues of consequence.
Although Trump introduced an executive order for a voluntary review of frontier AI models by the federal government earlier this year, the industry remains unregulated. Currently, enterprises deploy AI largely on their own terms and governance.
A Democrat-controlled Congress might eventually change that, Shay Boloor, chief market strategist at Futurum Equities, told CIO Dive in an email.
Data centers could see slower growth or more regulation under Democrats, passing more of the cost and pressure to build AI infrastructure to companies. And shifts in AI policy could influence CIO decision-making around model deployment and the pace of technology adoption.
“That creates a different policy environment around AI oversight and data center development, but still leaves Donald Trump in office with veto authority,” Boloor said. “The immediate market question becomes: What legislation can actually pass under divided government?”
AI policy
A sweeping federal AI law isn’t necessarily coming, but a change in party control does usually lead to a change in priorities, Stephen Cobb, a partner at Foley & Lardner who leads the law firm’s State Attorneys General practice.
This year has ushered in a more cautious era of AI due to growing costs, workforce challenges and an uncertain tech policy landscape. The changing makeup of Congress doesn’t change the legal positioning for companies already facing lawsuits or investigations, but it “creates an opportunity for progress,” Cobb said in an email.
The election results will only increase the likelihood of multistate efforts on tech policy, Cobb said. State attorneys general have been ahead on regulatory enforcement, according to Cobb, as they’ve been rolling out AI laws for years.
“The private sector has recognized that and is positioning itself accordingly,” he added.
Some AI labs have been trying to get ahead of regulations, such as OpenAI, which in May released its Frontier Governance Framework, a look into the company’s safety and security practices. The company said it planned to align with emerging state and global AI regulations, such as the EU AI Act.
“Members of Democratic leadership have been outspoken that they intend to flex regulatory oversight in Congress,” Cobb said. “Now that they have the gavels, we should see a heightened committee investigatory process, and AGs will certainly be paying attention.”
Data centers
The infrastructure that powers the compute for the growing AI industry had become largely unpopular with constituents, and voters sided against them in all but one of 41 ballot measures presented Tuesday.
The pushback comes as global AI spend is forecasted to reach $2.7 trillion this year, up nearly 50% from last year, as organizations further invest in the technology. Data center spending was the fastest-growing segment of IT spending, according to Gartner.
Texas, home to about one-quarter of planned U.S. data center capacity, elected a Democratic governor and senator Tuesday. Meaningful change in permitting, power costs or tax treatment could influence where and how quickly that infrastructure is built, Boloor said, but it is far from a stop to creating compute capabilities.
Governor-elect Gina Hinojosa has focused on pausing new development until additional safeguards are established such as giving communities more control, ending certain tax incentives and requiring developers to cover more of the costs of projects.
“That clearly makes development more expensive and potentially slower, but it is very different from saying AI infrastructure can no longer be built in Texas,” Boloor said.
The power shift toward Democrats could lead to a slower pace of data center construction, UBS said in a research note. Nearly one-third of expected data center additions for 2028 are at risk of being impacted, according to the report.
A moratorium on yet-to-be-built projects and tighter grid rules could reward companies and tech leaders that can solve their own power needs, Boloor said. The underlying demand for compute isn’t going anywhere, but the costs of building a functioning data center may now increase, making efficiency, power generation and infrastructure execution even more critical to companies relying on them.
“If future developments increasingly have to fund their own interconnection infrastructure or bring dedicated generation behind the meter, the advantage shifts toward developers and suppliers that provide power,” Boloor said.
