The Signal.
The signal, not the noise.
Today
The Signal.
Desks
The paper
Ideas · Essay

American AI faces a regulatory trap. China is positioned to sprint ahead.

Democratic leaders are calling for a slowdown, while Trump is pushing pro-AI policies. The real contest is not just technical; it is whether the U.S. regulates itself out of the race.
Monday, September 21, 2026

The question is now plain: is American AI doomed?

Tyler Cowen frames the fight as a collision between two forces. On one side, Democratic Party leaders including Barack Obama and Kamala Harris have echoed leading AI labs by calling for a slowdown. Senators Elizabeth Warren and Bernie Sanders want to pause or ban development altogether. On the other side, President Donald Trump is doubling down on pro-AI policies and saying the country that wins the AI race will win everything else besides.

That is the setup Cowen sees as a classic case of irresistible force meeting immovable object. The consequence matters far beyond Silicon Valley. If the United States responds to a strategic technology with a heavy regulatory hand, the incentive is obvious: investment slows, risk-taking falls, and the country surrenders speed to whoever keeps moving.

Cowen’s warning is not that the technology itself is a toy problem. It is that policy can decide the outcome. In his telling, the U.S. is slouching toward heavy regulation while China is poised to rush ahead. If that happens, Beijing could dominate everything from global arms sales to healthcare.

That is the part policymakers should not bury under slogans. The record is public: when the debate shifts from whether AI should exist to how fast it may be allowed to develop, the administrative state becomes a gatekeeper for a national industry. Gatekeepers do not build. They delay, deny, and demand permission slips.

Trump’s posture is the opposite. He is described here as doubling down on pro-AI policies, with a clear strategic claim: the winner of the AI race wins much more than prestige. For readers who care about free enterprise, that is the right question to ask. Not whether every model is harmless. Whether the United States intends to let caution become a self-inflicted tax on innovation.

The trade-off is real, and no serious person should pretend otherwise. AI carries obvious risks, and the calls for restraint are politically powerful. But the evidence in this piece points to a larger truth: if one country leans into caution while another leans into speed, the cautious country may hand over the future before the argument is settled.

That is the threat Cowen wants on the table. Not doom by machine, but doom by policy.

■

More from Ideas