Since mid-November, the White House has been signaling a desire to withdraw U.S. forces from Europe, shift security responsibility to the Europeans, and cut a deal with Russia that would restore U.S.-Russian economic ties while achieving “strategic stability” in Eastern Europe, in the words of last week’s National Security Strategy. Writing at The Scroll, I’ve argued that this strategy risks repeating the mistakes of the Obama and Biden administrations’ attempted Realignment with Iran. By positioning itself in Europe as a neutral mediator between NATO and Russia, rather than backing the former against the latter, Washington will strengthen Russian revisionism while weakening NATO’s ability to deter it. The result will be to create the instability this posture is attempting to avoid.

Defenders of the administration’s new approach, however, will argue that it is necessary for one big reason: China. Faced with a rising China, that is, an overstretched United States must pull its resources from Europe to concentrate on the primary threat in Asia, or the “Indo-Pacific,” to use the official jargon. This is the only way Washington can credibly deter Beijing.

More than a decade after President Barack Obama first announced the “pivot to Asia,” in other words, President Donald Trump is finally making it real.

Or is he? On Monday, President Trump made an unexpected announcement on Truth Social: The U.S. Commerce Department will begin allowing Nvidia to sell its H200 artificial intelligence chips in China “under conditions that allow for continued strong National Security.” According to the same post, 25% of the proceeds from the sales will be “paid to the United States of America.”

This strategy risks repeating the mistakes of the Obama and Biden administrations’ attempted Realignment with Iran.

What motivated the decision isn’t quite clear. Tech boosters in and around the administration, including AI czar David Sacks (a noted national-security dove) and Nvidia CEO Jensen Huang, have been aggressively lobbying the administration for months to allow for advanced chip sales to China. There was widespread speculation that Trump would announce approval for these sales at an October meeting with Chinese President Xi Jinping in October, but no such announcement occurred, with The Wall Street Journal later reporting that Secretary of State Marco Rubio and other China hawks had “torpedoed” Huang’s efforts. The decision to approve the H200s, but not the Blackwells, may represent an internal compromise between the China hawks and the tech boosters. But it may also simply represent the declining influence of figures like Rubio, as also evidenced by reporting around the Ukraine-Russia talks and in the dovish attitude toward China and Russia in the NSS.

But whatever the reasoning behind it, the decision met with almost universal condemnation from China hawks. While the H200 is not Nvidia’s most powerful chip (that would be the Blackwell), it is almost six times as powerful as the H20, the most advanced chip the United States currently allows to be sold in China. In a thread on X, Tim Fist of the Institute for Progress (IFP) noted that the H200 belongs to the generation of Hopper chips still used in most frontier AI research in the United States. China’s Huawei is not expected to produce a chip as powerful as the H200 until the end of 2027, and Chinese manufacturing bottlenecks mean that even those chips would not become widely available until 2028. “This move is giving China a bunch of advanced AI compute it wouldn’t otherwise have,” Fist wrote. Christopher Balding, a China analyst generally supportive of the administration’s strategy, wrote that the decision prioritizes “Chinese technological development of [that of] the United States” and will “actively [assist] the Chinese military and security state.” Like the administration’s TikTok decision, it is a “clear black mark on their record that threatens to derail their entire agenda.”

Read More

Elements of the U.S. government, moreover, appear to agree with that assessment. In what would prove to be a darkly comic example of bad timing, the U.S. Department of Justice unsealed a guilty plea in a case targeting a scheme to smuggle H200 and less-powerful H100 chips to China a few hours before Trump’s announcement. “These chips are the building blocks of AI superiority and are integral to modern military applications,” the DOJ said in its press release. “The country that controls these chips will control AI technology; the country that controls AI technology will control the future.”

Advocates of selling advanced AI chips to China argue that by offering Chinese firms American chips, Washington can slow the growth of China’s indigenous AI capabilities and ensure Beijing remains dependent on U.S. technology. This is a clever argument, but not a plausible one, IFP’s Santi Ruiz told Tablet. “The Chinese are desperate to develop their own indigenous supply anyway and are pouring huge resources into creating that demand signal and supply push,” he said. China’s “indigenous chip development is going to happen no matter what our export control regime is. The question is whether we want to close the gap for them by giving them the good stuff immediately.” Writing in The New York Times, David Sanger analogized the move to selling China F-35s because it is better to have them using American tech.

But OK, just because we’re offering the Chinese a leg up in the AI race doesn’t mean we’re not also preparing for a potential war, right? Well, maybe and maybe not. In the “maybe not” category, we have recent events with Japan. The NSS dropped language from previous administrations referring to China as a strategic competitor in Asia and called on the United States to shift the burden of deterring a Chinese invasion of Taiwan to its allies in the region. Indeed, one of the strategy’s main architects, Under Secretary of Defense for Policy Elbridge Colby, ruffled feathers in Tokyo over the summer by demanding that the Japanese clarify what role they would play in a crisis over the Taiwan Strait.

The Japanese did that last month, when Prime Minister Sanae Takaichi told parliament that a cross-strait invasion would constitute a “survival-threatening situation” that would permit the mobilization of Japan’s Self-Defense Forces. The Chinese responded with a torrent of threats and abuse, culminating in Chinese fighter jets locking their radar on Japanese F-16s near Okinawa over the weekend, a military provocation that drew a formal diplomatic protest from Japan.

In theory, Japan was doing just what the United States was asking of its allies—assuming diplomatic risks to clarify that it had its own skin in the game. Curiously, however, the Trump administration hesitated to back Tokyo. Asked last month about comments from China’s consul general in Osaka, who suggested Takaichi should be beheaded, Trump said merely that “a lot of allies aren’t our friend.” According to reporting in the Financial Times, the Japanese were privately expressing their “frustration” at the lack of diplomatic backing, which, prior to Tuesday, had amounted to nothing more than an X post from a deputy spokesperson and a vague reassurance from the U.S. ambassador to Japan. Finally, on Tuesday, a State Department spokesperson criticized the radar incident and said that “our commitment to our ally Japan is unwavering.” But the delay cannot have been reassuring, especially in the context of Washington signaling an overall desire for a détente.

A conciliatory tone toward China was also evident in recent remarks from Secretary of War Pete Hegseth. In May, Hegseth had said the Pentagon would be ready to fight China and win “decisively” in the event of an attempt on Taiwan. In comments over the weekend, however, he dropped this language and struck a far more dovish posture: “We are not trying to strangle China’s growth,” Hegseth said. “We are not trying to dominate or humiliate them, nor are we trying to change the status quo over Taiwan.” Instead, he said the United States was seeking “stable peace, fair trade, and respectful relations with China,” goals the War Department was pursuing with expanded dialogue with the People’s Liberation Army. In comments highlighted by Colby on X, Hegseth added that the U.S. goal in the Indo-Pacific was to achieve a “balance of power” with China.

In the spring, we often heard that Washington needed to reach a deal with Iran to quiet the Middle East and focus on strategic competition with China. Israel’s military campaign, followed by Operation Midnight Hammer, made that argument obsolete. Now the argument, recently advanced by figures like Donald Trump Jr.’s business partner Omeed Malik, is that the United States must pull off a “reverse Kissinger” with Russia, drawing it closer to Washington to enable us to focus on Beijing. Even as it presses forward with reducing its commitments to Europe, however, the United States appears to be signaling a lack of interest in confronting China, whether through economic or military means. This raises the obvious question of whether the much-advertised pivot to Asia is in fact a pivot to anything at all.