The Law as Weapon

Hugo Grotius was, at the time, on retainer to the Dutch East India Company. The pamphlet was extracted from a legal brief he had written to justify the Company's seizure of a Portuguese carrack in the Singapore Strait. The universal right of free navigation was, functionally, a rising commercial power's argument for why the incumbent's sea lanes were not the incumbent's.

England answered in 1635 with John Selden's Mare Clausum — the closed sea — asserting that sovereigns can and do own water, because England at that moment wanted its home waters and its fisheries under lock. Then England became the dominant naval power on Earth, and over the following century the Royal Navy became the world's great enforcer of freedom of the seas.

The doctrine flipped when the fleet flipped. It has flipped again since. The United States now holds the Mare Liberum brief and prosecutes it in the South China Sea.

I raise this because Foreign Affairs has published a careful, well-sourced account of Beijing mounting a sustained challenge to American legal hegemony — building extraterritorial statutes, training a generation of international lawyers, standing up parallel institutions, and using treaty architecture as an instrument of competition. Every fact in it appears to be true. The framing, that this represents an assault on a rules-based order, is the oldest sentence in the file. It has been argued from Lisbon, from Amsterdam, from London, and now from Washington.

i · extraterritoriality runs both directions, eventually

The specifics of what Beijing has built are worth stating plainly, because they are genuinely impressive and genuinely aggressive.

Since 2012, China has added extraterritorial provisions to more than twenty laws. It established a Foreign-Related Rule of Law Bureau inside the Ministry of Justice and a specialized international commercial court. It passed an Anti-Foreign Sanctions Law in 2021 and a Foreign Relations Law in 2023 explicitly framing legal instruments as tools of statecraft. In 2024 it designated fifty-one university-led partnerships binding law schools to government agencies, courts, firms, and arbitration bodies to mass-produce lawyers trained in foreign legal systems — a country that already had more law schools than the United States deciding it needed better ones. Party directives call for greater Chinese influence over the rules governing oceans, polar regions, cyberspace, outer space, nuclear safety, anticorruption, climate, and international finance. In 2026, State Council regulations created a mechanism to prohibit Chinese parties from complying with foreign assertions of jurisdiction Beijing deems improper; by May it had barred compliance with U.S. sanctions on five refiners accused of buying Iranian oil, and with the European Union's antisubsidy investigation into the security-equipment firm Nuctech.

That last instrument has a provenance worth naming. A blocking statute — a domestic law that makes it illegal for your own firms to obey another country's extraterritorial sanctions — is not a Chinese invention. It is the European Union's Blocking Regulation of 1996, written to shield European companies from Helms-Burton and the Iran and Libya Sanctions Act, which is to say written specifically to defend against American extraterritoriality. Canada and Mexico passed versions. So did the United Kingdom, decades earlier.

China's most-cited act of legal aggression in 2026 is built to a European blueprint drawn against Washington in 1996. Nobody in this story has invented anything; they are passing the same six statutes back and forth and renaming them according to who holds them.

Though I should be precise about what is copied, because the imprecision available here is the same one I just charged Foreign Affairs with. The architecture is shared: a domestic statute forbidding compliance with a foreign extraterritorial order. The control structure is not. Brussels' Regulation 2271/96 is a shield raised on behalf of private firms against a third country, subject to independent judicial review, and so famously under-enforced that European companies largely complied with the American sanctions anyway. Beijing's mechanism is an instruction issued to its own firms, applied at political discretion, reviewable by no institution outside the party-state, and used twice within months of taking effect. Same schematic, different hand on the switch — and the hand on the switch turns out to be the entire subject of this essay. Resemblance of form is not identity of thing. That rule cuts both directions or it is not a rule.

And then there is the part of the history that makes Chinese legal ambition the least surprising development of the century. Extraterritorial jurisdiction as a coercive instrument was demonstrated on China — the Treaty of Nanking in 1842, the Treaty of Wanghia in 1844, and the treaty-port system that followed, under which foreign nationals inside Chinese territory answered to foreign consular courts rather than Chinese ones. That regime lasted a hundred and one years. It was not fully abolished until 1943. The century that Beijing calls its humiliation was not administered primarily by armies. It was administered by law — by the specific proposition that jurisdiction is something a strong state exports and a weak state absorbs.

A state that learned what legal power is by being on the receiving end of it, and then spent forty years accumulating the economic mass to do something about it, is now writing extraterritorial statutes. There is no mystery here requiring analysis. This is the most predictable outcome in the archive.

ii · the incumbent broke the court first

The rules-based order narrative depends on an assumption that does not survive contact with the record: that the United States has been a participant in the legal architecture it authored.

The World Trade Organization's Appellate Body was an American achievement — binding, compulsory dispute settlement, the most legally serious adjudicative machinery in the history of international economic law, built substantially to American design in 1995. The United States then lost cases in it, objected to what it called judicial overreach, and beginning in 2017 blocked all new appointments. In December 2019 the body fell below quorum. Today all seven seats are empty. Two administrations of opposite parties maintained the blockade. The highest court of the trade order was disabled by the country that built it, on purpose, and the policy survived every change of government since — which is the tell that it is structural rather than partisan.

The United Nations Convention on the Law of the Sea was negotiated with heavy American involvement and has never been ratified by the United States Senate. Washington observes most of it as customary law and insists others treat it as binding. In 2016, an arbitral tribunal constituted under Annex VII of that convention ruled for the Philippines against China's nine-dash line in the South China Sea. China rejected the ruling and refused to participate. The United States demanded compliance with an award issued under a treaty it has declined for four decades to join.

The Rome Statute was signed by one American president and unsigned by the next. When the International Criminal Court later opened inquiries touching American and allied personnel, Washington sanctioned its prosecutor.

None of this makes the United States uniquely hypocritical. It makes it a hegemon. The consistent behavior of a dominant power in a legal order it authored is to treat the order as binding on others and advisory on itself — because that asymmetry is the return on having written it. Every previous holder of the position did the same. Britain enforced its anti-slavery patrols with a right of search over other nations' shipping that it would never have conceded over its own. Rome extended citizenship across the empire in 212, an act of legal generosity that also happened to widen the inheritance-tax base.

So legal hegemony feels like neutrality from the inside and like weaponry from the outside. That much is symmetric, and it explains why an American analyst writing about Chinese lawfare and a Chinese analyst writing about American legal hegemony are not disagreeing about facts. They are two observers reporting their position inside the same structure, accurately.

But the symmetry stops at the sensation, and the temptation is to let it run further than it goes. How an order feels is a function of where you stand in it. The order itself is not. There is a property that can be checked from outside every vantage point, and it is the only one that has ever mattered: does the architecture contain a binding third-party forum, with compulsory jurisdiction, that the strong actually submitted to — imperfectly, grudgingly, and for a while?

That question has an answer, and the answer is not a matter of perspective. It is why blocking appointments to the Appellate Body and rejecting the 2016 arbitral award are the same act, performed by opposite parties for opposite reasons. And it is why both are different in kind — not in degree — from a great power losing a case and complying with it, which has also happened, and more often than the cynical account allows.

iii · the cynic's conclusion is also wrong

Here is where this analysis usually terminates, and where it should not: therefore law is theater, only power is real.

That conclusion is false, and it is worth being precise about why, because it is exactly the conclusion each hegemon reaches at the moment it begins to lose, and reaching it is what converts a slow decline into a fast one. If the rules are only decoration, there is no cost to breaking them, and the first party to internalize that belief starts breaking them, and then the belief is true. It is a self-fulfilling prophecy with a detonation delay of about a decade.

Law does real work even when it is asymmetric. Predictability has value independent of fairness. A rule the strong break once in ten times is not the same thing as no rule: the breach is a scandal and the compliance is a civilization, both statements are true simultaneously, and the cynic's error is an unwillingness to hold two. The failure of the postwar order was never that it was fake. It was that its author held an exemption — and an exemption is a coherence maintained by suppressing the affected rather than including them, which is precisely the condition under which a legitimate order rots into a merely powerful one.

Which raises the actual question about the challenger, and it is not the question Foreign Affairs frames. The question is not whether China is contesting the order. It obviously is. The question is whether what Beijing is building would fix the exemption or relocate it.

Read the architecture and answer honestly. Extraterritorial statutes with domestic political control over their application. A blocking mechanism that lets the state instruct its firms which foreign law to disregard. A commercial court under a party-state judiciary. Directives to increase national influence over the rules governing every commons from the polar regions to outer space. This is not a proposal to make the order include the affected. It is an application to hold the exemption. It is the incumbent's own position, filed by a different applicant.

iv · who is actually standing in the blast radius

The parties genuinely destroyed by a two-hegemon legal contest are not the two hegemons. Both of them do fine. They are enormous, they have leverage that does not require a forum, and they can afford to route around any institution that rules against them.

The parties destroyed are everyone else — and specifically, the states for whom a binding forum was the only leverage they had ever possessed.

The WTO Appellate Body was, whatever its flaws, the one venue on Earth where a small economy could sue a large one and sometimes collect.

The sometimes is load-bearing, and the honest place to start is the case that failed. Antigua and Barbuda — population under a hundred thousand — took the United States to the WTO over its restrictions on cross-border online gambling. It won. It won the appeal. It was authorized to retaliate by suspending intellectual-property obligations. It never collected. The United States did not bring its measures into compliance, Antigua held a remedy it had no practical capacity to execute against an economy many thousand times its size, and the dispute sat unresolved for the better part of twenty years. That is the ceiling on what the forum could deliver to its smallest members, and the ceiling was low.

And then Costa Rica took the United States to the same body over restrictions on imported underwear, and won, and the restriction came off. Ecuador took the European Union over bananas and forced a regime open that had survived every bilateral approach attempted before it. A forum with a low ceiling and a real floor is a strange thing to defend, and I am defending it, because that floor is the only floor that has ever existed. It existed for twenty-four years. It is now gone — not killed by Beijing, but shut down by Washington, over objections that amounted to a court behaving like a court.

What replaces it is forum shopping. Chinese commercial courts against American district courts, Belt and Road arbitration against ICSID, competing sanctions regimes with competing blocking statutes so that a Brazilian or Indonesian or Nigerian firm cannot comply with one legal system without violating another. Every one of those choices gets adjudicated by relative leverage, which means the outcome is decided before the filing. For a mid-sized state, an order with one hegemon and a real court was better than an order with two hegemons and no court. That is an unpleasant sentence and it is arithmetically true.

Neither side's narrative mentions these parties. The American story is about defending a rules-based order. The Chinese story is about resisting hegemony. Both are stories in which the world's other one hundred ninety states appear as terrain.

And that sentence is still too generous to both of them, because it stops at states.

Go back to the party directives quoted earlier — the list of domains where Beijing intends greater influence over the rules. Oceans. Polar regions. Cyberspace. Outer space. Nuclear safety. Climate. International finance. Read the list again and notice what every item has in common: not one of them has a claimant. No state is the ocean's. There is no delegation for the orbital shell, no seat for the atmosphere, no standing for anyone who will be alive in 2100 and is not alive now. These are the domains where the consequences of the contest actually land, and they are structurally unrepresented in both architectures.

I have been applying a standard throughout this piece — that an order is legitimate to the extent it includes the affected rather than suppressing them, and that the American exemption is illegitimate because it excludes. Apply that honestly and it does not stop where I stopped it. Both orders answer the question who counts as affected? with states, and that shared answer is a far larger exemption than the American one. Older, deeper, and inherited entirely intact by whatever Beijing builds. The contest is over which power administers a set of commons inside a system where no commons can appear as a party.

Two hegemons fighting over the gavel is the story everyone is writing. That the courtroom has no chair for the plaintiff is the bigger one, and neither contestant has any incentive to notice.

v · the treaty port has moved

One further relocation, and it is not a prediction. It has already happened.

Extraterritoriality in 1842 required a physical port, a consular court, and a garrison — a patch of ground where foreign law displaced local law. The Treaty of Nanking needed geography. Nothing about the arrangement was subtle, which is part of why it eventually became intolerable enough to abolish.

Extraterritoriality in 2026 needs a server.

The American CLOUD Act of 2018 asserts that data held by a U.S.-controlled provider is reachable by U.S. legal process regardless of whose soil the drive sits on. The European Union's General Data Protection Regulation, at Article 3, reaches any firm anywhere on Earth that processes a European's data. China's Personal Information Protection Law and its data-export regime assert a mirrored claim in the other direction. None of the three is described by its authors as extraterritorial jurisdiction. All three are exactly that, and the ground they claim is not ground.

This is the treaty-port system reconstituted at the layer where value now sits. And it satisfies, precisely, the definition offered earlier for the humiliation: jurisdiction is something a strong state exports and a weak state absorbs. A Nigerian firm, an Indonesian bank, a Brazilian hospital — none of them chose which legal system reaches their records. That was settled by whose cloud they could afford, whose payment rail their customers use, whose models they run inference against. There may be four jurisdictions on Earth with meaningful sovereignty over the compute their own citizens increasingly think with. The remaining one hundred and ninety are in something close to the position China occupied in the century it now names.

The difference is that a treaty port was visible. You could photograph the consular court, walk its boundary, resent it accurately. Nobody can photograph a jurisdictional claim over a model's weights — and a regime nobody can see is a regime nobody abolishes.

vi · the prediction

Both powers will continue to denounce the other's lawfare while enacting substantively identical statutes, and both will be sincere, because from inside a legal architecture your own rules look like order and the other guy's look like aggression. Neither will notice the symmetry, because noticing it is expensive and nothing rewards it.

The instruments will keep migrating. The blocking statute was drafted in Brussels against Washington in 1996 and reached Beijing in 2021; extraterritorial jurisdiction went from the treaty ports to the Ministry of Justice in a hundred and eighty years; expect antitrust extraterritoriality and export-control regimes to complete the same circuit within the decade, each arriving described as unprecedented.

And in thirty or fifty years, whichever power holds the position will be arguing the case its rival is writing today. It will speak beautifully about universal principles and the common inheritance of humankind, and it will mean every word, and it will have a retainer.

Grotius did too. The seas are still contested — and so, now, is the substrate your own cognition runs on. The pamphlet is four hundred and seventeen years old and it still reads like a press release from last Tuesday.