Key facts
- Mike Gallagher, Palantir's head of defence, warned that the race for artificial intelligence makes it more tempting for Xi Jinping to seize Taiwan.
- Taiwan is home to TSMC, the world's top maker of advanced AI chips, and Nvidia has described the island as the epicentre of AI.
- Gallagher said Beijing could control or disable TSMC and then hold the global economy hostage.
- He spoke at a defence policy event in Indiana and has previously called Taiwan the 'single point of failure' for AI.
- China claims Taiwan as its territory and has not ruled out force. US intelligence has said an invasion would be extremely challenging and carry a high risk of failure.
- Palantir is a key Pentagon supplier, and its Maven AI system supports surveillance and targeting. NATO also uses it.
The AI race and Taiwan's strategic value
Taiwan's role in the global AI supply chain is central to Gallagher's warning. The island is home to TSMC, the world's top maker of advanced AI chips, and Nvidia has called it the epicentre of AI. That concentration of advanced manufacturing gives Taiwan an importance far beyond its size. The chips produced there sit inside data centres, military systems, smartphones, cars and the machines that are expected to power the next wave of artificial intelligence.
Gallagher argued that this dependence changes the strategic calculus in Beijing. If China were to seize Taiwan, it would either control TSMC or disable the company. Either outcome would give Beijing leverage over the economies and militaries that rely on advanced chips. In his view, that leverage could be used to hold the rest of the world economically hostage.
The warning comes as the United States and China compete for leadership in AI. Washington has restricted exports of advanced chips and chipmaking tools to China, while Beijing has poured money into domestic semiconductor production. The race is not only about commercial advantage. It is also about which country sets technical standards, controls critical supply chains and can field advanced AI in military systems.
Gallagher's warning
Gallagher, a former Republican congressman who led a House committee on China, joined Palantir in 2024. He has become one of the most vocal advocates for treating AI supply chains as a national security issue. At a defence policy event in Indiana on Wednesday, he said the logic of Xi Jinping sacrificing enormously to take Taiwan is getting more compelling.
His argument is that the AI race creates a closing window. If the United States and its allies pull far ahead in AI, China may conclude that its chances of catching up later are limited. That could make the near term more dangerous, not less. In Gallagher's words, if the West runs away with AI, now is the moment for Beijing to stop it, because China may not be able to catch up later on.
That logic does not mean an invasion is imminent. It means that the incentives around Taiwan are shifting as AI becomes more central to economic and military power. Gallagher's warning is aimed at policymakers who may see the AI race as a commercial contest rather than a strategic one. For him, the two are inseparable.
Why TSMC matters
TSMC is the world's largest contract chipmaker and a critical supplier for leading AI accelerators. Companies that design advanced chips rely on TSMC's manufacturing capacity, and the most advanced processors are difficult to produce without its specialised plants. This makes TSMC a chokepoint in the global technology system.
Taiwan's chip industry is also deeply integrated with the wider AI boom. Nvidia's graphics processing units, which are used to train and run large AI models, depend on advanced manufacturing. Data centre operators, cloud providers and governments are all competing for access to these chips. Any disruption to Taiwan's production would be felt across the technology sector and beyond.
Gallagher has previously called Taiwan the 'single point of failure' for AI. In a recent opinion piece, he urged the United States to arm the island heavily. His concern is that a small number of advanced manufacturing sites carry an outsized share of the risk. If those sites were damaged, captured or forced offline, the effects would not stay confined to Taiwan.
He said that if China takes Taiwan, it will control TSMC or disable TSMC. Beijing would then be able to hold the rest of the world economically hostage. That claim reflects a broader debate about how much of the global economy depends on a handful of advanced chip plants and how quickly alternative capacity can be built.
Beijing's position and US policy
China claims Taiwan as its territory and has not ruled out the use of force. At a White House summit with President Donald Trump in September, Xi urged Washington to oppose Taiwanese independence. The exchange underlined how Taiwan remains one of the most sensitive issues in the US-China relationship.
In May, Trump put a $14 billion arms package for Taiwan on hold, calling it a negotiating chip. The decision highlighted the tension between security commitments and diplomatic bargaining. Supporters of Taiwan argue that delaying arms deliveries weakens deterrence. Critics of the package say Washington should avoid provoking Beijing while seeking trade and diplomatic deals.
Gallagher's warning adds another layer to that debate. If AI is becoming the central technology of the century, then control of advanced chip manufacturing is not a peripheral issue. It is part of the balance of power. That framing could push Taiwan higher on the US policy agenda, even as Washington tries to manage its broader competition with China.
Military risk and intelligence assessments
US intelligence has played down the near-term risk of invasion. A March assessment found that an invasion would be extremely challenging and carry a high risk of failure. Crossing the Taiwan Strait, landing forces and sustaining them against a determined defence would be a massive military undertaking. China would also face the risk of economic retaliation and international isolation.
Still, China has stepped up coast guard patrols off Taiwan's east coast. These patrols are less escalatory than a military blockade or invasion, but they increase pressure on the island and complicate its defensive planning. They also signal that Beijing is willing to normalise a more assertive presence around Taiwan.
The intelligence assessment does not settle the question of long-term risk. Military balances can change, and political calculations can shift. Gallagher's argument is that the AI race could alter those calculations by making Taiwan seem more valuable and more urgent. If China believes that time is running out for its AI ambitions, the cost-benefit analysis in Beijing could become more volatile.
Palantir and military AI
Palantir is a key Pentagon supplier, and its Maven AI system helps US forces with surveillance and targeting. NATO also uses the system. The company's work sits at the intersection of software, data analysis and military operations, a field that is becoming more important as AI tools are integrated into command and control.
Gallagher's role at Palantir gives his warnings a commercial and operational dimension. He is not only talking about geopolitics in the abstract. He is speaking about the systems that militaries use to process information, identify targets and make decisions faster. The AI race is already shaping how armed forces organise themselves and what they prioritise.
That does not mean Palantir's technology decides the fate of Taiwan. It does mean that the company has a direct interest in how governments think about AI competition, export controls and defence readiness. Gallagher's public comments reflect that interest, but they also connect to a wider debate among defence officials and technology executives about the speed of AI adoption.
Self-improving AI and the nuclear comparison
Gallagher also pointed to talk in Silicon Valley of AI that can improve itself. If that is real, he said, it would be a game changer on par with nuclear weapons in the 1940s. The comparison is striking because nuclear weapons transformed military strategy, international relations and the balance of power in a very short period.
Self-improving AI remains a speculative idea, but it is taken seriously by some researchers and investors. The concern is that a system capable of improving its own capabilities could accelerate progress far beyond current expectations. If that happened, the country that controlled such technology could gain a decisive advantage.
For Gallagher, that possibility raises the stakes of the AI race and, by extension, the stakes around Taiwan. The island's chip industry is not the only factor in AI leadership, but it is a critical one. Control of advanced manufacturing, access to talent and the ability to scale compute all matter. So does the political stability of the region that produces the most advanced chips.
His warning is that the world may be entering a period in which technology competition and military risk reinforce each other. The AI race makes Taiwan more valuable. Taiwan's importance makes the AI race more dangerous. And the speed of AI development could compress the time available for policymakers to manage the risks.