Why We’re Skeptical About Congressional AI Regulations in 2027

This weekend, CEOs from a number of tech companies and large artificial intelligence (AI) firms began calling on Congress to pass new statutory regulations. The strategy of asking lawmakers for proactive industry regulations is a relatively new corporate concept that makes sense for a business facing massive public scrutiny and trying to avoid long-term regulatory uncertainty. 

In one recent example, social media platforms grew tremendously in an environment of limited federal regulation or liability, then faced retroactive government lawsuits, and eventually called for regulations on themselves. It is only logical that large AI firms would rather work with the government to set a statutory framework now and avoid repeating this disruptive cycle with lawmakers. 

Next week, President Trump and Speaker of the House Mike Johnson (R-LA) will hold a policy summit with leaders of major AI companies. While the issue of statutory regulations on AI technology development will surely continue to be a hot topic, there are several reasons we are skeptical that a regulatory framework can be signed into law this year or next.

President Trump Is Committed to a Deregulatory Space

As we’ve detailed before, President Trump and his staff at the White House see American leadership in AI technology as a legacy issue. The President understands the unpopularity of the industry but nonetheless has been committed to “winning the AI race” since the early days of this administration. We do not anticipate that this will change. That means for the next two years, Congress will try to pass regulations against a backdrop of an uncooperative White House and veto threats. Sweeping new regulations will need super-majority support to become law. 

Congress Does Not Have Enough Information to Move

Congress simply doesn’t have the knowledge base to quickly regulate AI programs, codes, algorithms, applications, and models. As Speaker Johnson said earlier today, “if everyone was in a room and wanted to regulate AI, no one would know what to do.” We agree. There will have to be months of hearings, staff education, engagement with AI companies, lobbying, drafting, mark-ups, and bicameral agreement. This process will be highly complicated in an environment with a Democratic majority that will be hard pressed to work with President Trump.

Conservatives Coalescing Around the Fear of Regulatory Capture

Skepticism around the motives and intentions of large AI companies among lawmakers is real. Those concerns have not been alleviated this weekend just because many of those companies are calling for more federal regulations. In fact, conservatives are already advancing the notion that large firms want to establish costly regulations that would stifle innovation or create barriers to entry that make it impossible for new companies or competitors to pop up. This argument is quickly emerging in conservative think tank circles and demonstrates a Catch-22 for AI regulation – any new laws supported by current AI leaders will be viewed with suspicion by lawmakers and their constituents. 

Politicizing AI Regulations for the Midterms Makes Bill Next Year More Difficult

This weekend’s hype around AI regulations is already being perceived by Republicans as a political campaign stunt by Democrats ahead of the midterms. Polls show AI regulations are popular, and that fact isn’t lost on Democrats. After President Trump defended data centers last week, Democrats saw a chance to take the popular side of a political wedge issue. Republicans on the Hill are already saying that recent concern about AI is a cynical campaign move by the Democrats as opposed to legitimate policymaking. This perception within the GOP is real and will slow efforts on a bill that will need 60 votes in the Senate next year. 

The China Race

The notion of “beating China” is a clear goal for lawmakers and a natural argument to pump the breaks on regulations. Senator Steve Daines (R-MT), who sits on the Senate Foreign Relations Committee, said just today, “It is imperative that the United States wins this race. And for those who assert we should somehow slow down is nonsense. This is about continuing to move as quickly as possible because we are in a race. The bronze or silver medal is not good enough.” This sums up a national security perspective that is shared among many Republicans and materially complicates any legislation to regulate models, labs, or technology development. 

Regulatory Measures Have Not Even Been Crafted

While Congress doesn’t have the sophistication to regulate AI models or labs right now, it certainly can grasp how to regulate other aspects of the industry. There are already bills in existence restricting companies from doing business with companies in adversarial countries, restricting outbound investment, banning Chinese robots and AI models, or forcing data centers to pay for infrastructure and energy costs. These types of bills that regulate the periphery of the AI industry are the types of legislation that could move in the current environment. But regulations on the technical aspects of AI development are not yet ready for prime time.