The Question Just Got Personal
Days ago, Xi Jinping looked Donald Trump in the eye in Beijing and warned that mishandling Taiwan would put the US-China relationship in "great jeopardy." Not a statement. Not a press release. A face-to-face warning between the two most powerful men on Earth.
That makes the question unavoidable: if China moves on Taiwan, should the United States fight for it?
This is not an academic exercise. It is the single most consequential strategic decision the US may face this decade. It deserves an honest answer, not a slogan. Here is the real case on both sides — and a real position at the end.
First, Frame the Question Correctly
Most of the debate is sloppy because it collapses three different questions into one.
Question one: should the US help Taiwan defend itself with weapons, training, and intelligence? Question two: should the US commit, explicitly and in advance, to direct military intervention if China invades? Question three: should the US maintain deliberate ambiguity about what it would do?
These are not the same question. You can answer yes to the first and no to the second. The entire US policy of "strategic ambiguity" for the past 45 years has been exactly that — arm Taiwan, train Taiwan, but never say for certain whether American forces would intervene. Keep both Beijing and Taipei guessing.
Hold those three questions separate. They have different answers.
The Case For Defending Taiwan
The semiconductor argument. Taiwan's TSMC fabricates roughly 90% of the world's most advanced semiconductors. Every iPhone, every Nvidia GPU, every frontier AI system depends on chips made within artillery range of mainland China. If China controls Taiwan, China controls the supply of the most strategically critical resource of the 21st century. That is not a regional shift. That is the end of American technological supremacy.
The credibility argument. The entire American alliance system runs on credibility. Japan, South Korea, the Philippines, Australia, NATO — all of them stay aligned with Washington because they believe the US keeps its commitments. Abandon Taiwan and every ally recalculates overnight. Japan and South Korea would have to seriously consider their own nuclear weapons. The hub-and-spoke alliance system that has underwritten 80 years of relative peace would unravel.
The geography argument. Taiwan sits at the center of the "first island chain." It is the cork in the bottle. As long as Taiwan is friendly or neutral, the Chinese navy is bottled into its near seas. Lose Taiwan and the PLA Navy has open, permanent access to the broader Pacific. The military map of Asia is redrawn.
The precedent argument. If China successfully takes Taiwan by force and the world accepts it, the rule that borders cannot be changed by conquest is dead. That rule is the foundation of the post-1945 order. Its collapse invites every revisionist power to test its own neighbors.
The deterrence argument. Appeasement of an aggressor historically invites more aggression, not less. The 1930s are the cliché reference for a reason. A China that takes Taiwan cheaply is a China emboldened to push further.
The Case Against — Or For Serious Caution
The nuclear argument. This is the one that separates Taiwan from every other US commitment. China is a nuclear power with a rapidly expanding arsenal. A direct US-China war over Taiwan carries genuine risk of nuclear escalation. Ukraine, for all its stakes, never carried that risk because the US was never directly fighting Russia. Taiwan would mean American forces directly killing Chinese forces. That is a category of risk that should terrify any honest analyst.
The military reality. Taiwan is 100 miles off China's coast and 7,000 miles from the United States. China has spent 25 years and enormous resources building anti-access and area-denial capability designed for one specific purpose: making US intervention near Taiwan prohibitively costly. They have built the missiles, the submarines, and the targeting systems specifically to sink American carriers. The home-field advantage is not close.
The economic mutual destruction argument. US-China trade runs into the hundreds of billions of dollars per year, and the two economies are deeply entangled through supply chains. A war does not just damage two militaries. It detonates the global economy — a depression-scale event affecting every country on the planet, allied or not.
The overextension argument. This is the realist case, most associated with John Mearsheimer. The US is already committed in the Middle East, still backstopping Ukraine, and now staring at Taiwan. A great power that spreads itself across three simultaneous theaters invites failure in all of them. The realist position is not isolationism — it is ruthless prioritization. Pick the theater that matters most and concentrate there.
The Wargame Nobody Wants to Read Closely
The Center for Strategic and International Studies ran the most rigorous public wargame of a 2026 Chinese invasion of Taiwan. They ran it 24 times. The result deserves to be read slowly.
What the CSIS wargame found: "The United States might win a pyrrhic victory, suffering more in the long run than the 'defeated' Chinese." The US lost dozens of ships, hundreds of aircraft, and roughly 3,200 troops in the first three weeks alone. Taiwan's navy was destroyed in most scenarios. — CSIS, "The First Battle of the Next War"
Read the full ledger. The US and its allies: dozens of ships sunk, hundreds of aircraft lost, thousands dead in three weeks. Taiwan: navy destroyed, economy wrecked, an island left without electricity and basic services. China: roughly 10,000 troops killed, 155 aircraft and 138 major ships lost, 30,000 soldiers captured.
The wargame's central finding is the part that matters most. Taiwan usually survives as an independent entity — but the cost to the United States is so severe that Washington's global position is damaged for years. You can "win" this war and still lose the decade that follows.
The Chip Argument Cuts Both Ways
The semiconductor case is the strongest argument for defending Taiwan. It is also more complicated than it first appears.
The "Silicon Shield" theory holds that China will not invade because it needs TSMC's output too badly to destroy it. There is something to that. But it cuts the other way too: in any actual war, TSMC's fabs are among the first things destroyed — by Chinese missiles, by Taiwanese sabotage to deny them to the invader, or simply by losing the power and water and skilled workforce that advanced fabrication requires. The chips do not survive the war you fight to save them.
This is the uncomfortable truth. A Taiwan whose value you can only protect by threatening nuclear war is not a strategic asset. It is a strategic liability. Which points directly at the policy that should be happening regardless of any war decision: aggressively diversifying advanced chip production away from Taiwan. Apple is reportedly already doing exactly this, evaluating Intel and Samsung as fabrication backups — a shift we analyzed in depth here. Every advanced fab built in Arizona, Ohio, or Japan reduces the leverage Taiwan's chips give Beijing and lowers the stakes of the whole confrontation.
The Real Answer
The honest position is not "yes, fight" or "no, abandon." It is a specific strategy with three parts.
One: arm Taiwan to the teeth for its own defense. The answer to question one is an emphatic yes. Taiwan should be a porcupine — sea mines, mobile anti-ship missiles, drones, hardened command systems, a population trained in resistance. The goal is "deterrence by denial": make a successful invasion look impossible to Beijing's own planners. This is cheap relative to a war and it does not put the US on an automatic escalation ladder.
Two: keep direct US intervention deliberately ambiguous. The answer to question two is no — do not issue an explicit, advance, unconditional commitment to send American forces. Strategic ambiguity has kept the peace across the strait for 45 years. It deters China (you might intervene) while restraining Taiwan (you might not, so do not declare independence and drag the US into a war it did not choose). An explicit guarantee removes the restraint on Taipei and hands Beijing a clear target to plan against.
Three: diversify the chips on a wartime timeline. The semiconductor concentration is the real vulnerability. It should be treated as a national emergency independent of any China decision. The faster advanced fabrication exists outside Taiwan, the lower the stakes of the entire confrontation and the more freedom of action the US actually has.
Put simply: make Taiwan impossible to swallow, stay ambiguous about whether you would personally intervene, and remove the prize that makes the island worth the gamble. That is deterrence without an automatic path to nuclear war. It is the position a serious realist and a serious China hawk can both accept — which is exactly why it is the right one.
The Bottom Line
Xi warned Trump about Taiwan to his face this week because Beijing wants Washington thinking about the cost. The wargames say that cost is staggering even in victory. Anyone who tells you the answer is simple — fight no matter what, or walk away no matter what — is not being honest with you.
The defensible answer is the harder one. Help Taiwan make itself indigestible. Keep your own hand hidden. And urgently dismantle the chip dependency that turned a small island into the most dangerous flashpoint on Earth. Deterrence is not the promise of war. It is making the war look unwinnable to the only person whose opinion matters — the one deciding whether to start it.
