Earlier this month, Senator Josh Hawley declared war. In a speech delivered at American Compass’ annual gala—a black-tie affair attended by some of the most influential conservative thinkers in DC—he cast the unfettered development of AI as an existential threat to the country. Sounding less like a U.S. Senator and more like an Old Testament prophet, Hawley warned of the bitter fruits of AI maximalism. A ‘K-shaped’ economy in which elites thrive while the masses are immiserated. The concentration of an unprecedented amount of power in an incredibly small number of private hands, accountable to nobody except their shareholders. The obliteration of the wage labor economy, and the elimination of the ordinary American’s ability to make a living by his hands or his brain. The endangerment of our children and the irreversible corruption of public morals.

“In deciding how to govern this technology,” he boomed, “we are not merely writing policy. We are renewing—or surrendering—the moral basis of our life together. The American covenant does not keep itself. Every generation must swear it again. Ours will swear it, or break it, over artificial intelligence.”

Bracing stuff. But not entirely unexpected from Missouri’s senior Senator, who has recently emerged as the Right’s most prominent and compelling AI skeptic. Over the past year or so, he’s made this an increasingly important part of his political brand, chairing Senate hearings on the technology’s harms, introducing bills to restrict everything from data centers’ power consumption to kids’ use of chatbots, and hammering the greed and recklessness of AI developers at every opportunity.

Hawley, it should be noted, is an outlier among his party. Indeed, arguably the biggest booster of AI today is the Trump White House, which under the influence of figures like David Sacks and Marc Andreessen, has done everything in its power to eliminate constraints on the technology and encourage faster development and wider dissemination. Many—most?—of Hawley’s Congressional confrères are probably closer to the Sacks position, and the threat of a presidential smackdown surely cows some of the other skeptics.

Happily, there are signs the broader movement is starting to realize the true dimensions of the threat. As a Politico story pronounced the other day, “a GOP revolt over AI is taking shape.” We catch occasional glimpses of this in DC—the defeat of federal preemption during last year’s reconciliation process, for example—but most of the action is happening at the state level, where Republican governors and legislators are increasingly targeting data centers and championing AI safeguards. Figures like Florida’s Ron DeSantis and Utah’s Spencer Cox are stepping in where the feds will not.

And little wonder: poll after poll shows that ordinary Americans are profoundly concerned about what this technology could do to them, their kids, and their country. 57% of registered voters believe the risks of AI outweigh the benefits. 65% think AI will destroy more jobs than it creates. 71% oppose the construction of new data centers in their communities. And 81% support strong government guardrails to mitigate potential harms.

But if the great GOP AI revolt is to succeed, it will need to be something more than a series of reflexive responses to developments already underway. What is called for is a positive, disciplined approach, one that soberly yet urgently addresses both the risks and the promises of AI.

An approach, in other words, like the one Hawley outlined in his speech.

As with any good plan, this vision is anchored by three overarching priorities.

The first is the protection of working Americans. “We must regulate artificial intelligence to ensure that it aids the worker and does not replace him,” Hawley said, sketching an array of pro-labor policy interventions: restrictions on labor-obviating applications like self-driving cars, the extension to workers of rights over AI in their companies, and licensing rules to protect functions and positions that require a human touch.

The second is the establishment of what might be called AI subsidiarity. The Senator specifically criticized the buildout of data centers, which has proceeded with unprecedented rapidity and minimal transparency—often over the objections of the communities in which they are situated. But the principle is broader: sovereignty over AI must be vested in the American people, and decisions about the technology should be made at the local level whenever possible.

The third priority is the protection of society’s most vulnerable—especially our children. Following a spate of incidents in which chatbots assisted or encouraged children in harming themselves, Hawley introduced a bill to restrict minors’ access to such ‘AI companions’. That bill, known as the GUARD Act, cleared the Judiciary Committee with unanimous support, and appears to have kindled a new interest in legislative checks on AI-driven harms to kids.

Taken together, these priorities point the way to a genuinely conservative blueprint for AI policy. They avoid the radical overcorrections favored by the hard left—universal basic income, nationalization—while still meaningfully addressing the most profound technological development since the Industrial Revolution. For the difficult but undeniable fact is that unregulated AI constitutes a clear threat to authentic human flourishing. As Pope Leo XIV observed in Magnifica Humanitas, AI presents us with the age-old temptation of Babel in a new form: “the idolatry of profit that sacrifices the weak, a uniformity that neutralizes differences, and the pretense that a single language — even a digital one — can translate everything, including the mystery of the person, into data and performance.”

It is incumbent on us to reject this temptation, and to use public power to ensure that AI is aligned with the public good. To protect jobs and ensure that economic change actually benefits workers. To preserve local agency in the face of unprecedented corporate clout. To shield society’s most vulnerable from new and rapidly developing dangers.

No amount of GDP growth is worth losing our republic. Conservatives should therefore follow Hawley’s lead and build out a vigorous, comprehensive program to ensure that AI works for Americans—not the other way around.