Trump takes 15% cut from Nvidia China sales while cancelling Taiwan defence meetings
Former officials warn systematic concessions undermine technological competition with Beijing
In April, Donald Trump banned advanced semiconductor exports to China, calling them a national security threat. This week, he reversed course and began collecting a 15% commission on those same sales. The president's explanation was refreshingly blunt: "I want you to pay us as a country something, because I'm giving you a release."
The Nvidia deal represents more than presidential whiplash. It signals a shift from treating export controls as security tools to operating them as profit centres. For former Trump administration officials who designed America's technological competition strategy, this transformation represents a dangerous precedent that plays directly into Beijing's hands.
The timing tells the story. After Nvidia CEO Jensen Huang attended a $1 million-per-head dinner at Mar-a-Lago, the White House abandoned plans for tighter chip restrictions. Instead, Trump negotiated what he proudly described as haggling Huang down from a 20% government cut to 15%. Meanwhile, Chinese negotiators sent 75 officials to recent Stockholm trade talks while America dispatched just 15—a disparity that reveals who wants this deal more.
The profitable reversal
Trump's revenue-sharing arrangement breaks new ground in strategic competition. The deal allows Nvidia and AMD to sell artificial intelligence chips worth an estimated $35 billion annually to Chinese companies, generating roughly $5 billion for the US government. Treasury Secretary Scott Bessent calls this "the model and the beta test" for expansion to other industries.
The president characterises the H20 chip as "obsolete" and "an old chip that China already has", but technical experts flatly disagree. "These H20s are still state of the art," says CSIS analyst Emily Kennedy, with "elements that make them extremely sophisticated and valuable". The chips powered China's DeepSeek AI model, which stunned Silicon Valley earlier this year by achieving advanced capabilities at a fraction of expected costs.
This technical gap between Trump's dismissive characterisation and expert assessment reveals either profound misunderstanding or deliberate political spin. Either possibility raises uncomfortable questions about how America makes strategic technology decisions.
The arrangement transforms decades of export control philosophy. Since the Cold War, technology restrictions served clear strategic purposes: deny adversaries critical capabilities that enhance military power. Now they become licensing schemes where security policies carry price tags. The precedent is stark—every future export control becomes a potential revenue opportunity rather than an immutable security barrier.
Expert warnings ignored
Matt Pottinger, Trump's former deputy national security adviser, led 19 other policy professionals in urging the administration to reject H20 sales to China. Their letter called the decision "a strategic misstep that endangers the United States' economic and military edge in artificial intelligence". The administration proceeded anyway.
"This deal suggests that what gets banned or permitted is not being driven by careful calculations about Chinese military power—but rather on political whim and personalist politics," warns Jennifer Lind, a Dartmouth international relations expert. The concern extends beyond individual decisions to systemic breakdown of strategic coherence.
China has spent years learning to exploit Trump's dealmaking obsession. The revenue-sharing model gives Beijing exactly what it wants: normalisation of American concessions through transactional relationships. Each deal establishes precedent for the next, steadily eroding the principle that some technologies remain off-limits regardless of price.
The broader pattern is unmistakable. Export controls originally aimed to maintain American technological advantages have become bargaining chips in personal negotiations. When security policies become subject to individual dealmaking, patient adversaries hold systematic advantages over democratic leaders facing electoral pressures.
China's patient extraction strategy
The chip concession forms part of a carefully orchestrated Chinese campaign to extract systematic American retreats. Beijing's approach combines economic pressure with diplomatic patience, understanding that Trump values visible deals over strategic consistency.
China secured perhaps its most significant victory on Taiwan. The United States blocked Taiwan President Lai Ching-te from visiting New York after Beijing objected, with sources citing Trump administration fears that Chinese displeasure would derail trade negotiations. Earlier, America cancelled meetings between Taiwan's defence minister and senior US officials immediately following a Trump-Xi phone call.
Each concession builds toward the next. When Trump revealed that Xi Jinping promised never to invade Taiwan "as long as you're president", the Chinese leader added a crucial caveat: "I am very patient, and China is very patient". This wasn't merely about Taiwan timing—it was strategic communication about China's long-term approach to extracting American concessions through patient pressure.
Beijing deploys economic leverage systematically. After Trump's tariffs, China completely halted rare earth exports—materials critical for advanced technology production—then used their resumption as negotiating currency. The message was clear: China possesses escalation advantages in economic confrontation that America struggles to match.
The American Enterprise Institute warns that these precedents create cascading vulnerabilities: "The PRC could interpret the decision as an indication of the US's willingness to concede on issues on which the PRC places special emphasis". Once America demonstrates willingness to trade core strategic positions for immediate gains, every future negotiation begins from that baseline.
Strategic implications of transactional security
The revenue-sharing model raises fundamental questions about American strategic coherence when core policies become profit opportunities. Export controls remain effective only whilst America possesses technologies China needs. But Chinese firms are rapidly developing workarounds that could make American restrictions irrelevant within years.
Since export controls began, China initially suffered significant disruption, with semiconductor output falling 17% in early 2023. However, restrictions also triggered massive Chinese investment in domestic alternatives. Beijing now pursues what experts call "an all-out, government-backed effort to improve self-sufficiency in all aspects of semiconductor design and production".
The competitive mathematics are unforgiving. Semiconductor leadership requires enormous ongoing investment in research and development. If export controls reduce American firms' revenues through lost Chinese sales, they undermine the very innovation capacity the controls aimed to protect. CSIS research warns of potential "death spirals" where reduced competitiveness leads to less R&D investment, further reducing competitiveness.
Trump's approach reflects what former officials recognise as fundamental misunderstanding of strategic competition with authoritarian systems. As Matt Pottinger observes: "We're not dealing with a fellow democracy here, where there's a baseline of trust that people are working for the good of their people. Leninist systems work for whichever party wants to maintain power".
The 15% arrangement may generate billions in government revenue, but it establishes that American strategic technologies are negotiable commodities rather than tools of national power. For allies assessing American reliability, and for China calculating future concessions, the signal is unmistakable: sufficient economic incentives can override security considerations.
Whether this represents shrewd dealmaking or strategic capitulation will depend on factors beyond Trump's control—how China uses these technologies, what precedent revenue-sharing sets for other critical sectors, and whether American technological leadership can survive systematic conversion of security policies into personal profit opportunities.
The fundamental question is whether America can maintain strategic advantages against patient adversaries willing to pay incrementally higher prices for systematic erosion of the policies designed to contain them.