J.D. Tuccille: Trump's plan preempt regulation of AI is welcome

In 2024, Colorado passed legislation to, supposedly, protect consumers from discrimination by artificial intelligence (AI). Arguably, the law does more to require the incorporation of diversity, equity, and inclusion ideology (DEI) into the growing technology than to protect people from anything. That has the White House worried that state and local micromanagers will cripple innovation with politicized mandates. In response, U.S. President Donald Trump issued an executive order to standardize and limit AI regulation across the country. It’s a promising approach, but it needs firmer support from congressional action.

The Colorado law imposes on those who develop and deploy AI a “duty to avoid algorithmic discrimination” defined as “any condition in which the use of an artificial intelligence system results in an unlawful differential treatment or impact that disfavours an individual or group of individuals on the basis of” a laundry list of protected racial, sexual, and other statuses.

The law includes significant compliance and reporting requirements that could easily induce AI developers to build all sorts of preferences into the system to avoid offending any member of a protected class. The mandates also create substantial red tape hurdles to entering the field — so much for starting a new project in your garage, unless you build an addition for lawyers.

Concerned that they might have created a bureaucratic monster, state lawmakers considered several revisions. The only one that passed, however, did nothing more than delay implementation from Feb. 1, 2026, to June 30, 2026. Colorado isn’t a big deal by itself, but plenty of people hoping to take advantage of the new technology worry that politicians across the country will strangle it in its crib.

“We don’t like seeing blue states trying to insert their woke ideology in AI models, and we really want to try and stop that,” commented David Sacks, a tech investor who advises the Trump administration.

The White House initially favoured legislation to head off state AI regulation. President Trump warned, “We MUST have one Federal Standard instead of a patchwork of 50 State Regulatory Regimes. If we don’t, then China will easily catch us in the AI race.” But Congress isn’t especially good at getting things done these days (not necessarily a bad characteristic for lawmakers) and that didn’t happen. Instead, the White House issued an executive order.

Signed Dec. 11, the order stipulates that “AI companies must be free to innovate without cumbersome regulation” and that “State-by-State regulation by definition creates a patchwork of 50 different regulatory regimes that makes compliance more challenging, particularly for start-ups.” It commits the administration to work with Congress on a “minimally burdensome national standard” that bars conflicting state rules. It also sets up a litigation task force to challenge state AI laws. A separate task force will review laws for provisions that, among other failings, violate the First Amendment.

The executive order also conditions federal broadband funding and other discretionary grants to the states on the absence of onerous AI regulations. It directs the Federal Trade Commission (FTC) to investigate whether legislation like that adopted by Colorado which “require alterations to the truthful outputs of AI models” runs afoul of federal prohibitions on engaging in deceptive acts or practices.

The usual suspects immediately complained that “the Trump-Sacks executive order proves that the White House only listens to powerful Big Tech CEOs who fund ballrooms rather than the everyday people they pretend to serve,” as Sacha Haworth, Executive Director of The Tech Oversight Project, put it. Her group wants government control of pretty much all technology,

But some thoughtful people had a very different take.

Jessica Melugin, director of the Competitive Enterprise Institute’s Center for Technology and Innovation had previously called for exactly such federal preemption, writing in May, “it’s absurd on principle that AI technologies wouldn’t obviously qualify as interstate commerce and, therefore, be a matter for federal instead of state authorities.”

After Trump signed the Dec. 11 order, Melugin responded, “we need federal preemption of most state AI regulation in order to successfully compete with China in the race to lead AI. Much of AI commercial activity is clearly interstate commerce. A patchwork of 50 different regulatory regimes places our entrepreneurs and firms at an unfair advantage globally. Congress should codify this approach.”

Neil Chilson, head of AI policy at the Abundance Institute, is equally encouraged by the White House order. A former FTC official himself, he commented, “the FTC provision remains of great interest to me, both because of my background and because it is very creative.” He added that the final language of the order “excludes specific areas of otherwise lawful state law from a preemption recommendation. This neutralizes the non-stop rhetoric that this is about a total federal takeover.”

In a detailed analysis, Chilson writes that while the order “is not a silver bullet” it “raises the cost of the worst forms of state AI regulation, creates institutional pressure to test their legality, and clearly signals that the status quo of 50 competing AI regimes is unacceptable for a technology that operates at national and global scale.”

The biggest hurdle the AI executive order faces is that it isn’t legislation. While often vastly powerful in their scope, such orders are essentially just memos to executive branch personnel dictating priorities and interpretations of existing law; they can’t actually change the law. To the president’s credit, his executive order recognizes its limitations. It specifies that the White House will prepare a legislative recommendation to preempt excessively intrusive state laws restricting AI.

But the legislative proposal will have to get through Congress. That’s not a guarantee at a time when confidence in free markets and innovation is weak and suspicion of technology is on the rise. Polling — much of it tendentiously worded, to be clear — finds majority support for state regulation of AI among both Democrats and Republicans.

But heavy regulation of AI by any level of government won’t guarantee anybody’s safety. It will just ensure that development of promising technology will take place in friendlier jurisdictions. Trump’s AI executive order isn’t perfect, but it’s a step in the direction of heading off excessive red tape.

National Post