Trump is right: AI guardrails are a gift to China
By Tom Beckwith · Reporting from Washington ·
In the corridors of the State Department, we were taught that credibility is a currency earned in the boring years and spent in the crises.
In the corridors of the State Department, we were taught that credibility is a currency earned in the boring years and spent in the crises. Right now, the United States is attempting to spend its AI credibility on a whim, and the currency is plummeting.
The current spat between President Donald Trump and the CEOs of the AI frontier is not a debate about ethics; it is a collision of interests. On one side, we have Dario Amodei, the CEO of Anthropic, who recently published a blog post arguing that we must "pace the frontier" to avoid a catastrophe. He is joined by the likes of Sam Altman and Elon Musk, who have suddenly discovered a conscience regarding the existential risks of their own products. On the other, we have a President who views "guardrails" as a synonym for "handcuffs." As reported by CNBC and The Guardian, Trump has dismissed these calls for regulation, asserting on Truth Social that the only guardrail the U.S. needs is a "STRONG AND SMART (High IQ!) PRESIDENT."
The Altruism of the Billionaire's Hedge
Let us be clear: when a man whose company is valued at nearly a trillion dollars begs the government to slow him down, he is not acting out of a sudden devotion to humanity. He is managing his balance sheet. White House AI czar David Sacks hit the mark when he suggested that the motivation to slow down is not altruistic, but a shield against "massive product-liability exposure." If an AI enables a catastrophic cyberattack, the company wants a regulatory framework to share the blame—or a government-mandated slowdown to ensure the product doesn't break before the IPO.
The "existential risk" argument—the claim by former Anthropic researchers that AI could "kill us all by the end of the decade"—is the strongest case the regulators have. They argue that we are building a god we cannot control and that a global moratorium is the only rational response. But this is a luxury of the leading. To pause is to assume that your competitors will also pause. In the real world, interests trump altruism. This is the same dynamic we saw during the Net Neutrality Repeal, where the priority shifted from theoretical consumer protections to the raw capacity for industry growth and global competitiveness. The "safety" plea is a Trojan Horse, as Vice-President JD Vance noted, designed to protect the incumbents from the volatility of their own creations.
Sputnik Logic in a Silicon Age
The President’s rhetoric is, as usual, a blunt instrument, but his geopolitical intuition is correct. He describes a "sick conspiracy" against AI and data centers—a phrase CNA notes he uses to frame the debate—and argues that the only party benefiting from American hesitation is China.
This is the 1957 Sputnik Crisis all over again. When the Soviets launched that first satellite, the West didn't form a committee to discuss the "safety" of orbital mechanics or the ethical implications of space debris. They recognized a technological gap as a strategic vulnerability. The Sputnik Crisis taught us that in a zero-sum race, the only way to manage risk is to win the race first. The mechanism is identical here: the framing of a breakthrough as a geopolitical necessity renders domestic regulation a strategic handicap.
We have seen this playbook before. During the Trump-China Trade War, regulatory restrictions were framed as weaknesses that Beijing would exploit. The current insistence that "whoever wins AI wins" is not theater; it is the only metric that matters in a capital that views the world through the lens of capabilities. House Speaker Mike Johnson is right to warn that an emergency session to regulate AI would be a gift to China. In a world of competing hegemonies, the first nation to achieve a decisive AI advantage doesn't just win a market; they rewrite the rules of global power.
The High Cost of Playing the Angel
The tragedy here is that the President is treating a strategic imperative like a personal grudge. By slamming Dario Amodei as a "perfect little angel" and threatening "TREASONISTS" and "LEAKERS," Trump is applying the same aggressive posture he used in the TikTok Ban Attempt—targeting individuals to signal strength. But this approach is a relic of the COVID-19 Pandemic Response, where intuition was substituted for institutional frameworks with mixed results.
The path forward is predictable. With midterms on November 3, Speaker Johnson will ensure no meaningful AI legislation reaches the floor. The "race to China" narrative will swallow any nuanced discussion of safety. Meanwhile, Anthropic will likely find its planned 2026 IPO complicated by a President who views its CEO as a conspirator.
The CEOs are playing a game of liability management, and the President is playing a game of dominance. Between the two, the realist must side with the latter. You cannot regulate a technology that your primary adversary is developing without constraints. The only effective guardrail in a global arms race is the lead. If we pace ourselves into a second-place finish, we won't be worrying about the "risks" of AI; we will be living under the rules written by the people who didn't stop to ask if it was safe.