What the Alien Tort Statute Lets an American Court Do

A foreign victim of torture can walk into a federal courthouse in the United States and file a lawsuit over abuse that happened entirely in another country. That sentence surprises most readers, and the surprise is the reason the Alien Tort Statute matters. The statute is thirty-three words long, passed by the First Congress in 1789, and for most of American history it did nothing at all. Then, in 1980, a Paraguayan family used it to sue a Paraguayan torturer in a Brooklyn courtroom, and the statute woke up. What followed was the most ambitious experiment in the history of human rights law: an attempt to turn American courts into a forum for the world’s worst wrongs. The experiment lasted four decades, produced a line of Supreme Court decisions that narrowed the opening with each pass, and ended with the Court’s declaration that judges may not invent new claims under the statute at all.

The thesis of this article is that the Alien Tort Statute’s two-century arc proves that American courts may hear the world’s worst cases only when Congress has clearly said so, because creating a remedy for a foreign wrong is an act of foreign policy, and judges were never given that power. Every great case in the statute’s history has been a fight over the same question: who gets to turn the key, Congress or the judge. The answer, delivered across six Supreme Court decisions, is Congress. Understanding why requires walking the whole arc, from a diplomatic embarrassment in 1784 to a technology company’s surveillance contracts two centuries later, because the doctrine only makes sense as a story about the separation of powers wearing the clothes of human rights.

The reader who finishes this article will be able to explain exactly who can sue whom in American courts for abuses committed abroad, why the answer narrowed with each decade, and what remedies remain for victims now that the statute’s door has nearly closed. That is a map worth having, because the question underneath it never goes away: when the courts where the abuse happened cannot or will not act, where does the victim go.

The Thirty-Three Words, Read Closely

The Alien Tort Statute, codified at 28 U.S.C. section 1350, provides that the district courts shall have original jurisdiction of any civil action by an alien for a tort only, committed in violation of the law of nations or a treaty of the United States. Read it again slowly, because the entire doctrine lives inside the grammar. The statute grants jurisdiction, which is the power to hear a case. It does not, on its face, grant a cause of action, which is the legal right to demand a remedy. That distinction, between the power to hear and the right to recover, is the hinge on which the next two hundred years turn, and the Supreme Court’s final word on the statute is essentially a long meditation on what that hinge can bear.

Three elements sit inside the sentence, and each one has generated its own body of law. First, the plaintiff must be an alien, meaning a noncitizen of the United States. An American victim of overseas abuse cannot use this statute, a limitation that surprises readers and that Congress later addressed with a separate law. Second, the claim must sound in tort, the civil law of wrongs like assault, battery, and false imprisonment, rather than in contract or some other field. Third, the tort must violate either the law of nations, the eighteenth century’s term for what is now called customary international law, or a treaty of the United States. The treaty prong has produced almost no litigation of consequence, because treaties that create privately enforceable rights are rare and courts construe them narrowly. Nearly the whole history is about the law of nations prong.

Two features of the text deserve emphasis because later courts made so much of them. The statute says nothing about where the tort occurred. It does not say “committed within the United States” or “committed abroad.” That silence became, in the 2013 Kiobel decision, the ground for applying the presumption against extraterritoriality, the rule that federal statutes are presumed to govern only inside American borders unless Congress says otherwise. And the statute says nothing about who the defendant may be. It does not say “committed by an American” or “committed by a foreigner.” That silence became the ground for the corporate liability fights of the 2010s, which asked whether a business entity could be hauled into court for abuses committed by others. A thirty-three word sentence, written before the steam engine, was asked to answer questions about multinational supply chains and internet surveillance infrastructure. The strain shows, and the strain is the story.

Why 1789 Needed a Law Like This

The First Congress did not pass the Alien Tort Statute out of humanitarian idealism. It passed it out of embarrassment and fear, which are the more reliable legislators. In May 1784, a French adventurer named Charles Julian de Longchamps assaulted the French consul general, Francois Barbe-Marbois, on a Philadelphia street, and later threatened him at his residence. Under the Articles of Confederation, the Continental Congress had no court in which to prosecute an assault on a foreign diplomat, and Pennsylvania’s courts handled the matter only after awkward delay. The French government protested. For a fragile confederation seeking credit, trade, and military alliances in Europe, the inability to punish insults to foreign envoys was not a minor procedural gap. It was a signal to every European chancellery that the United States could not guarantee the safety of their representatives.

A second incident sharpened the point. In 1787, a Dutch diplomat’s household in New York was the scene of an intrusion that again exposed the absence of a federal forum. The pattern was clear: the law of nations, which in the eighteenth century governed the treatment of ambassadors, the safety of foreign merchants, and the conduct of piracy, imposed obligations on the new nation, but the nation had built no court competent to enforce them. William Casto, the Texas Tech historian who has written the definitive account of the statute’s origins, argues in his 2004 study that the provision was a jurisdictional gap-filler, designed to assure foreign powers that violations of their nationals’ rights under the law of nations would find a federal court. The purpose was diplomatic credibility, not global justice.

This origin matters because it fixes the statute’s original meaning in a narrow frame. The First Congress worried about three classic violations of the law of nations as understood in 1789: offenses against ambassadors, violations of safe conduct, and piracy. William Blackstone’s Commentaries on the Laws of England, published between 1765 and 1769 and widely read in America, listed exactly these three as the principal offenses against the law of nations. The statute’s drafters, Oliver Ellsworth chief among them, almost certainly had that triad in mind. Nothing in the debates, sparse as they are, suggests that anyone in 1789 imagined the law being used by a Paraguayan farmer against a Paraguayan police chief for torture committed in Asuncion. The original understanding was modest, state-centered, and diplomatic. Everything expansive that came later had to be built on top of that modest foundation, and the Supreme Court eventually decided that the foundation could not carry the weight.

A famous early case shows how naturally American courts of that era handled foreign claims, and it makes a useful contrast with what the statute later became. In the Amistad proceedings of 1839 to 1841, kidnapped Africans who had seized their slave ship were tried in American courts, and the Supreme Court, in an opinion by Justice Joseph Story, ordered them freed. The full reconstruction of that case shows a young republic’s highest court adjudicating the rights of foreign nationals caught in an international wrong. But Amistad was not an Alien Tort Statute case. It was about property, treaties, and the status of free persons under American and Spanish law. The distinction is worth holding onto: American courts have always heard cases with foreign elements. The question the statute raised was whether they could hear a foreigner’s tort claim against another foreigner for conduct abroad, with no American interest at stake beyond the abstract interest in upholding international law.

What the Law of Nations Meant to the First Congress

What did the First Congress mean by the law of nations?

The First Congress meant the body of customs and principles that governed relations among sovereign states and the conduct of individuals under their protection, principally the safety of ambassadors, the sanctity of safe conducts, and the suppression of piracy, as described by Blackstone and the continental jurists the founders read.

The longer answer is where the doctrine’s deepest argument lives. The “law of nations” in 1789 was not a code. It was a set of customs, treatises, and shared understandings drawn from the practices of European states, and its content was contested even then. Emer de Vattel’s The Law of Nations, published in 1758, was the most influential single work, and the founders cited it constantly. Vattel described a system in which sovereigns owed each other duties of noninterference and in which certain acts, like piracy, made the perpetrator an enemy of all mankind, hostis humani generis, subject to capture and punishment by any nation. That pirate analogy later became the human rights movement’s favorite precedent: if any nation could punish a pirate in 1789, why could not any nation punish a torturer in 1980. The analogy is rhetorically powerful and legally slippery, because pirates operated outside all sovereign protection on the high seas, while modern torturers usually operate inside a sovereign’s territory with its approval.

The originalist debate over the statute turns on how fixed the phrase “law of nations” was meant to be. One camp, associated with Casto’s historical work, holds that the phrase referred to a closed or at least stable set of eighteenth-century norms, and that the statute therefore authorizes only claims analogous to the three Blackstone offenses. The other camp, which carried the day in the lower courts during the 1980s and 1990s, holds that the phrase was meant to be dynamic, referring to whatever the law of nations prohibits at the time of the suit, so that as international law grew to condemn torture, genocide, and crimes against humanity, the statute grew with it. Judge Irving Kaufman’s opinion in Filartiga adopted the dynamic view explicitly, writing that courts must interpret international law as it exists at the time of decision rather than as it stood in 1789.

The Supreme Court’s answer in Sosa v. Alvarez-Machain, decided in 2004, was a compromise that pleased neither camp fully. Justice David Souter, writing for the majority, held that the statute was meant to have a dynamic element, that Congress in 1789 expected courts to recognize new claims as the law of nations evolved, but that the evolution had to be disciplined. A modern claim would be actionable, Souter wrote, only if it rested on a norm of international character accepted by the civilized world and defined with a specificity comparable to the three eighteenth-century paradigms. That sentence, the Sosa test, became the gate through which every later claim had to pass, and very few passed. The compromise preserved the dynamic reading in theory while making it nearly unusable in practice, which is a pattern the Court would repeat.

The Hundred and Ninety-One Year Silence

For one hundred and ninety-one years, from 1789 to 1980, the Alien Tort Statute produced essentially no litigation. Two reported decisions in the twentieth century invoked it, and neither went anywhere. The reasons for the silence are as instructive as the noise that followed, because they reveal what had to change in the world before the statute could be used.

First, the nineteenth century’s international law was a law of states, not of persons. Individuals were objects of international law, not subjects of it. A Paraguayan farmer tortured by Paraguayan police had no standing in the conceptual universe of Vattel or his successors, because the law of nations governed what states owed each other, and how a state treated its own citizens was its own business. The idea that international law could regulate a government’s treatment of its own people, the core premise of modern human rights law, did not exist in any enforceable form until after the Second World War.

Second, American courts of the nineteenth and early twentieth centuries had no occasion to stretch the statute. The paradigmatic cases it was written for, assaults on ambassadors and piracy, were handled through criminal prosecution and admiralty jurisdiction. Civil tort suits by aliens for international law violations simply did not occur to litigants, because the legal imagination of the era did not contain the category.

Third, and most decisively, the human rights revolution had not happened. The Universal Declaration of Human Rights was adopted in 1948. The International Covenant on Civil and Political Rights opened for signature in 1966. The Convention Against Torture was adopted in 1984. Each of these instruments helped crystallize the idea that certain abuses, torture above all, violate norms binding on all states regardless of treaty ratification. Without that crystallization, there was no “law of nations” prohibition for a court to apply to a modern abuse. The statute slept because the law it referenced had not yet been written.

What woke it was a convergence in the late 1970s: a maturing body of international human rights law, a generation of American lawyers trained in the civil rights movement’s litigation strategies, and a set of horrifying facts that demanded a forum. The Center for Constitutional Rights, founded in 1966, had spent the 1970s developing the model of impact litigation, the deliberate selection of cases to change law rather than merely to compensate clients. When the Filartiga family arrived in the United States with their story, the lawyers saw not just a tragedy but a vehicle. The statute had been waiting for someone to notice it, and the human rights bar finally did.

Filartiga v. Pena-Irala: The Revival

The facts of Filartiga v. Pena-Irala are the kind that make a legal doctrine feel inevitable in hindsight, which is why they deserve to be stated plainly. Joel Filartiga was a Paraguayan physician and opponent of the dictatorship of Alfredo Stroessner. In March 1976, Paraguayan police seized Filartiga’s seventeen-year-old son, Joelito, tortured him to death, and displayed his body to his mother as a warning. Americo Norberto Pena-Irala, the police inspector-general of Asuncion, was alleged to have supervised the torture. The Filartiga family later learned that Pena-Irala had entered the United States on a visitor’s visa and was living in Brooklyn, New York. In 1978, having had Pena-Irala served with process while he awaited deportation, the family filed suit in the Eastern District of New York, invoking the Alien Tort Statute.

The district court dismissed the case. Judge Eugene Nickerson held that the law of nations, as understood for jurisdictional purposes, governed relations between states and did not reach a state’s torture of its own citizens. The Second Circuit reversed in 1980, in an opinion by Judge Irving Kaufman that became the founding document of modern transnational human rights litigation. Kaufman’s reasoning ran in three steps. First, he held that the law of nations must be interpreted as it exists at the time of decision, not as it stood in 1789, citing the evolution of international law since the Second World War. Second, he found that the prohibition of official torture had become a norm of customary international law, pointing to the Universal Declaration of Human Rights, the International Covenant on Civil and Political Rights, the European Convention on Human Rights, and a stack of General Assembly declarations. Third, he concluded that the torturer, like the pirate before him, had become an enemy of all mankind, so that any nation’s courts could hear the claim.

The opinion’s most quoted passage declares that the torturer has become like the pirate and slave trader before him, an enemy of all mankind, and that international law therefore gives every state the right to exercise jurisdiction over him. The passage is eleven words of analogy doing the work of a legal system. Its power came from making the unfamiliar feel traditional: the court was not inventing a new power but extending an old one. The Second Circuit remanded the case, and in 1984 the district court entered a default judgment of more than ten million dollars against Pena-Irala, who had long since returned to Paraguay and never paid. The judgment was unenforceable in any practical sense, and that did not matter to the lawyers. The point had never been compensation. The point was the precedent, and the precedent held.

Filartiga’s significance is difficult to overstate and easy to misstate. It did not hold that American courts could hear any human rights claim from anywhere. It held that official torture, committed by a state actor, violated a sufficiently definite and universal norm to support jurisdiction under the statute. The case’s limits were built into its reasoning: the defendant was served inside the United States, the norm invoked was among the most settled in international law, and the court emphasized the consensus of the world’s nations. Later litigants would try to stretch each of those limits, and later courts would spend decades deciding how far the stretching could go.

The Expansion Years: Kadic, Marcos, Unocal

The decade and a half after Filartiga was the statute’s heroic age, when lower federal courts steadily widened the opening Kaufman had cut. Three lines of expansion mattered most, and each one later became a target.

The first line extended liability to private actors. In Kadic v. Karadzic, decided by the Second Circuit in 1995, Bosnian Muslim and Croat women sued Radovan Karadzic, the Bosnian Serb leader, for genocide, war crimes, and crimes against humanity committed during the Bosnian war. Karadzic argued that the law of nations bound only states, and that as a private faction leader he could not violate it. The court disagreed, holding that certain norms, genocide and war crimes foremost among them, bind private individuals as well as state officials. The reasoning drew on the Nuremberg precedent, where the international military tribunal had convicted private persons and organizational leaders for crimes against humanity. The criminal side of that history, the trials that established individual responsibility under international law, is reconstructed in the Nuremberg trials account, and the civil ATS cases of the 1990s were its direct doctrinal heirs, translating criminal liability into civil damages.

The second line extended the statute to the architects of kleptocracy. In the Marcos litigation, thousands of Filipino victims of the Ferdinand Marcos regime sued the former dictator’s estate in Hawaii, where Marcos had fled after the 1986 People Power Revolution. The Ninth Circuit allowed the claims, and in 1994 a jury awarded nearly two billion dollars, then the largest human rights judgment in history. The Marcos cases mattered because they showed the statute reaching the very top of a foreign government, and because the litigation dragged on for decades through asset recovery proceedings, demonstrating both the symbolic power and the practical limits of transnational judgments.

The third line, and the most consequential for what followed, extended liability to corporations. In Doe v. Unocal, filed in 1996, Burmese villagers sued the American oil company Unocal for its alleged complicity in forced labor, murder, and rape committed by the Burmese military during construction of the Yadana gas pipeline. The theory was aiding and abetting: Unocal had not committed the abuses itself but had allegedly benefited from and facilitated them. The Ninth Circuit allowed the case to proceed toward trial in 2002, and Unocal settled in 2005 on the eve of trial, reportedly for tens of millions of dollars. The settlement meant no appellate ruling on corporate aiding-and-abetting liability, but the case terrified multinational legal departments. If an American company could be sued in American courts for what a foreign army did along a foreign pipeline, the statute had become a global regulatory instrument, and the business community began organizing for a Supreme Court correction.

By the turn of the century, the lower courts had recognized ATS claims for torture, extrajudicial killing, genocide, war crimes, crimes against humanity, and, in some circuits, forced labor and environmental harms. The statute that slept for 191 years was suddenly the busiest corner of human rights law. The expansion rested on an assumption that the Supreme Court had never tested: that federal judges possessed a common-law power to recognize new international norms as actionable. That assumption was about to meet its examiners.

The First Backlash: Tel-Oren and Judge Bork

The expansion did not go unchallenged, and the challenge came from inside the judiciary. In Tel-Oren v. Libyan Arab Republic, decided by the D.C. Circuit in 1984, survivors of a Palestine Liberation Organization terrorist attack in Israel sued Libya, the PLO, and others under the ATS. The court dismissed the case, but the three judges wrote three separate concurrences, and one of them became the most influential losing argument in the statute’s history.

Judge Robert Bork’s concurrence argued that the Alien Tort Statute granted jurisdiction only and created no cause of action, and that international law itself did not supply one either. In Bork’s view, Filartiga had been wrongly decided: the statute opened the courthouse door, but no substantive law walked through it, because the law of nations consisted of norms addressed to states, not of private rights enforceable by individuals in court. Without a cause of action created by Congress, the court had nothing to adjudicate. Bork warned that letting judges fashion causes of action from international law would transfer foreign policy from the elected branches to the judiciary, a constitutional objection that would echo through every later case.

Bork’s concurrence lost in 1984. It won slowly, over the next four decades, as its core claims were adopted first by scholars, then by dissenting justices, then by majorities. Curtis Bradley, the Duke law professor who became the leading academic skeptic of ATS expansion, built his scholarship on Bork’s foundation, arguing in a series of articles that the Filartiga line rested on a misreading of both the statute and the constitutional allocation of foreign affairs power. Bradley’s position, stated most fully in his work on the Charming Betsy canon and extraterritoriality, was that courts should not use vague international norms to displace the political branches’ control of foreign relations. The business lobby, led by the U.S. Chamber of Commerce and the National Foreign Trade Council, translated the academic argument into amicus briefs, warning that ATS suits punished American companies for operating in difficult countries and handed foreign policy to plaintiffs’ lawyers.

The backlash also had a democratic theory behind it that deserves a fair statement. International law, unlike a statute passed by Congress, is not made by any legislature accountable to American voters. Treaties require Senate consent, but customary international law emerges from the practices and statements of foreign governments, international organizations, and scholars. Letting judges enforce customary international law as if it were federal common law meant, in the skeptics’ view, letting the world legislate for the United States without the United States voting. That objection never fully answered the human rights side’s response, which was that torture and genocide are not policy disagreements but universal prohibitions, but it gave the Supreme Court a constitutional vocabulary for narrowing the statute, and the Court used it.

Sosa v. Alvarez-Machain: The Supreme Court Intervenes

The case that brought the statute to the Supreme Court began with a kidnapping arranged by the American government. Humberto Alvarez-Machain, a Mexican physician, was accused of participating in the 1985 torture and murder of Enrique Camarena, an agent of the U.S. Drug Enforcement Administration. In 1990, the DEA arranged for Mexican nationals, including Jose Francisco Sosa, to seize Alvarez-Machain in Mexico and deliver him to the United States for trial. Alvarez-Machain was eventually acquitted, returned to Mexico, and sued Sosa and the United States, claiming among other things that his arbitrary detention violated the law of nations and was therefore actionable under the Alien Tort Statute.

The Ninth Circuit agreed that arbitrary detention stated an ATS claim. The Supreme Court reversed unanimously on that point in 2004, in an opinion by Justice David Souter that became the statute’s new charter. Sosa v. Alvarez-Machain, 542 U.S. 692, is the most important ATS decision ever issued, because it settled the question Bork had raised twenty years earlier: does the statute create causes of action, or must courts find them elsewhere.

Souter’s answer was characteristically careful. The statute, he wrote, is strictly jurisdictional. It opens the courthouse door but does not itself create the claim. However, the First Congress had enacted it against the background of a common law that recognized a small set of international norms as giving rise to private suits, the Blackstone triad of offenses against ambassadors, violations of safe conduct, and piracy. Congress therefore expected courts to recognize new claims as the law of nations evolved, but only within tight discipline. Souter then announced the test: a modern claim is actionable only if it rests on a norm of international character accepted by the civilized world and defined with a specificity comparable to the features of the eighteenth-century paradigms.

The test had two visible effects. First, it killed Alvarez-Machain’s own claim, because the prohibition of “arbitrary detention,” while real in human rights instruments, was not defined with anything like the specificity of piracy. A brief, cross-border seizure for trial did not violate a sufficiently definite norm. Second, it preserved Filartiga in principle while making its replication difficult. Torture, Souter implied, might still qualify, because the prohibition of official torture is specific and universal. But each new candidate norm would face skeptical judicial scrutiny, and the scrutiny would consider not only the norm’s definition but the practical consequences of recognizing it, including the foreign policy fallout.

Souter added a further caution that later majorities would sharpen into a blade. He wrote that courts should consider whether the political branches had acted, that they should be wary of claims with no American connection, and that the judiciary’s role in foreign affairs counseled restraint. The opinion thus contained both the statute’s survival and the seeds of its confinement. Human rights advocates celebrated that the Court had not closed the door. Skeptics noted that Souter had described a door so narrow that few would fit through it. Both readings were correct, and the next two decades proved the skeptics righter.

The Sosa Test, Step by Step

The Sosa test sounds abstract until a court applies it, so it is worth walking through the mechanics the way a district judge would. A plaintiff files an ATS complaint alleging, for example, that a foreign official subjected the plaintiff to prolonged arbitrary detention. The judge’s first question is whether the norm invoked, here the prohibition of arbitrary detention, is defined with the specificity of piracy. Piracy had a precise common-law definition: robbery on the high seas, with settled elements. Arbitrary detention, by contrast, shades from a traffic stop into a disappearance, and human rights instruments define it in general language. Under Sosa, that generality is fatal. The judge dismisses.

A second plaintiff alleges torture by police, with medical evidence and a consistent account. The judge asks the same question and gets a different answer. The Convention Against Torture, adopted in 1984 and ratified by more than one hundred and seventy states, defines torture with precision: the intentional infliction of severe pain or suffering, by or with the acquiescence of a public official, for purposes such as obtaining information or punishment. That definition is specific, the prohibition is universal, and the norm has the character of jus cogens, the peremptory law from which no state may derogate. The claim survives the Sosa gate. This is why, even after Sosa, torture remained the paradigm ATS claim, and why litigants kept framing new abuses as species of torture.

The test’s third component is the one most readers miss: Souter instructed courts to consider the practical consequences of recognizing a claim, including whether doing so would interfere with the political branches’ conduct of foreign affairs. This is not a legal test in the ordinary sense. It is an invitation to judicial statesmanship, and different judges answered it differently. Some treated it as a soft presumption against novel claims. Others, particularly after Kiobel, treated it as a reason to dismiss any case touching a country with which the United States had sensitive relations. The consequence consideration meant that two identical complaints could come out differently depending on the defendant country’s relationship with Washington, which sits uneasily with the idea of law as a system of rules but accurately describes how the doctrine worked.

The Sosa test thus functioned as a funnel: wide enough at the top to admit the historic core of torture, extrajudicial killing, and genocide, and narrow enough at the bottom to exclude everything else. Litigants responded the way litigants always do, by characterizing their claims to fit the funnel. Environmental harms became cruel and degrading treatment. Labor abuses became slavery. Forced displacement became crimes against humanity. Some of these characterizations succeeded in sympathetic district courts. Almost none survived appellate review after 2013, because the appellate courts increasingly read Sosa’s specificity requirement as demanding near-identity with an eighteenth-century paradigm rather than family resemblance. The funnel kept narrowing, and the narrowing was the point.

Jurisdiction Is Not a Cause of Action

What is the difference between jurisdiction and a cause of action?

Jurisdiction is a court’s power to hear a dispute, while a cause of action is the substantive legal right to demand a remedy for a wrong. A court can have jurisdiction over a case, the power to decide it, and still dismiss because no law gives the plaintiff a right to win.

The distinction is the most misunderstood concept in the entire ATS debate, and it is worth mastering because everything else depends on it. When Congress gives federal courts jurisdiction over “all civil actions arising under the Constitution, laws, or treaties of the United States,” the famous federal question jurisdiction of 28 U.S.C. section 1331, it is not creating a right to sue. The right to sue comes from the underlying law: the civil rights statutes, the securities laws, the Constitution itself as interpreted. Jurisdiction is the door. The cause of action is what walks through it.

The Alien Tort Statute’s text grants only the door. “The district courts shall have original jurisdiction of any civil action by an alien for a tort only, committed in violation of the law of nations or a treaty of the United States.” Compare that with the Torture Victim Protection Act, which provides that an individual who subjects another to torture or extrajudicial killing “shall, in a civil action, be liable for damages.” The TVPA’s language creates liability. The ATS’s language creates only a forum. Bork’s entire concurrence in Tel-Oren was an elaboration of this grammatical observation, and Sosa adopted it as the statute’s authoritative reading.

Why does the distinction matter so much? Because causes of action, unlike jurisdiction, implicate the separation of powers directly. Creating a new right to sue is legislation. When a court recognizes a new ATS claim, it is effectively writing a small statute: it defines the prohibited conduct, identifies the liable parties, and sets the remedy. Sosa permitted courts to do this interstitial lawmaking only for norms as settled as the eighteenth-century paradigms, on the theory that Congress in 1789 had authorized that much. Everything beyond it required Congress to act. And Congress, with one significant exception, never did. The distinction between jurisdiction and cause of action is therefore not a technicality. It is the constitutional reason the statute’s expansion stopped.

The Torture Victim Protection Act of 1991

Congress did act once, and the result is the second great instrument of transnational human rights litigation. The Torture Victim Protection Act, signed by President George H. W. Bush in 1992 after passage in 1991, provides an express cause of action for torture and extrajudicial killing committed under actual or apparent authority, or color of law, of any foreign nation. The statute’s sponsor in the Senate, Arlen Specter of Pennsylvania, framed it as a response to Filartiga: the courts had found a door, and Congress would now build a proper entrance.

The TVPA differs from the ATS in every structural respect, and the differences illuminate both statutes. First, the TVPA creates liability explicitly, so the Bork objection has no purchase. A court applying the TVPA is enforcing Congress’s command, not inventing its own. Second, the TVPA is limited to two abuses, torture and extrajudicial killing, each defined in the statute with care. Torture tracks the Convention Against Torture’s definition. Extrajudicial killing means a deliberate killing not authorized by a previous judgment of a regularly constituted court, excluding lawful combat deaths. The specificity that Sosa demanded of ATS claims is built into the TVPA’s text.

Third, the TVPA requires exhaustion of adequate and available remedies in the place where the conduct occurred. A plaintiff must first try the courts of the country where the torture happened, and may come to America only if those remedies are ineffective or unavailable. The exhaustion requirement answers the sovereignty objection: the United States does not displace foreign courts but supplements them where they have failed. In practice, exhaustion is rarely a barrier, because the countries that produce TVPA cases are usually the ones whose courts cannot hear them, but the requirement structures the litigation and gives judges a disciplined inquiry.

Fourth, the TVPA carries a ten-year statute of limitations, with equitable tolling for extraordinary circumstances, which has allowed cases decades after the abuse. The ATS has no limitations period of its own, and courts borrowed the TVPA’s ten years for ATS claims as well, a borrowing that shows how the two statutes grew intertwined.

The TVPA’s most important judicial gloss came in Mohamad v. Palestinian Authority, decided by the Supreme Court in 2012. The Court held unanimously, in an opinion by Justice Sonia Sotomayor, that the TVPA’s cause of action runs only against natural persons, not organizations. The statute says “an individual” who subjects another to torture shall be liable, and the Court held that “individual” means a human being, not a corporation or other entity. The decision closed the TVPA to suits against companies and groups, a limitation that would later be echoed in the ATS corporate cases and that made the choice of defendant the central strategic question in transnational litigation.

Two Doors, Different Keys

A reader who has followed this far can now see the litigation landscape as lawyers saw it in the early 2000s: two doors into American courts for foreign abuses, each with its own key. The ATS door admitted aliens only, covered any tort violating the law of nations, and required no exhaustion, but its cause of action had to be judicially recognized under Sosa’s demanding test. The TVPA door admitted any plaintiff, including Americans, but covered only torture and extrajudicial killing, required exhaustion, and ran only against natural persons after Mohamad.

Plaintiffs’ lawyers routinely pleaded both, using the ATS for the broad customary-law claims and the TVPA for the solid statutory ones. The pairing was strategic: if the court balked at recognizing a novel ATS norm, the TVPA’s express cause of action for torture might still carry the case. Defendants attacked both, arguing that the TVPA’s specificity implied that Congress had occupied the field and that courts should not use the ATS to go beyond what the TVPA allowed. That argument, that the later, narrower statute implicitly limited the earlier, broader one, gained force with each passing year and found its fullest expression in the Cisco litigation.

The two-door system also produced a persistent puzzle about American plaintiffs. An American tortured abroad could not use the ATS, because the statute requires an alien plaintiff, but could use the TVPA. A foreigner tortured abroad could use either. The asymmetry reflected the statutes’ different theories: the ATS was about giving foreigners a forum for violations of universal norms, while the TVPA was about giving torture victims, of any nationality, a remedy against their torturers. The distinction made sense on paper and generated endless motion practice in reality, as defendants probed each plaintiff’s citizenship and each claim’s characterization.

By the mid-2000s, the system had a stable shape: a narrow but real set of claims, a sophisticated plaintiffs’ bar, an organized defense bar, and a Supreme Court watching from above. The stability did not last. The Court took Kiobel in 2012, and the extraterritoriality question, which Souter had flagged but not decided, moved to center stage.

Kiobel and the Presumption Against Extraterritoriality

The facts of Kiobel v. Royal Dutch Petroleum read like a test case designed in a seminar room, which in a sense they were. Nigerian plaintiffs, members of the Ogoni people of the Niger Delta, alleged that Royal Dutch Petroleum and Shell Transport and Trading, through their Nigerian subsidiary, had aided and abetted the Nigerian military government in committing extrajudicial killings, torture, and crimes against humanity during the suppression of Ogoni protests in the 1990s. The abuses occurred in Nigeria. The defendants were Dutch and British corporations. The only connection to the United States was that the plaintiffs had been granted asylum here and that the companies did business here. The Second Circuit dismissed the case, holding that the law of nations does not recognize corporate liability. The Supreme Court took the case to decide the corporate question, heard argument on it, and then, in an unusual move, ordered reargument on a different question entirely: whether the ATS applies extraterritorially at all.

The reargument signaled where the Chief Justice’s mind was going. In April 2013, the Court held unanimously in result, though divided in reasoning, that the presumption against extraterritoriality applies to the Alien Tort Statute. Chief Justice John Roberts, writing for the five-justice majority, reasoned that the presumption is a canon of statutory construction reflecting the principle that American law governs domestically but does not rule the world. Because the ATS’s text contains no clear indication that Congress meant it to reach conduct in foreign territory, the presumption applies, and claims must be dismissed unless they “touch and concern” American territory with sufficient force to displace the presumption.

The phrase “touch and concern” came from the Court’s earlier extraterritoriality cases, and Roberts gave it no precise definition, which guaranteed a decade of litigation over its meaning. What was clear was the direction: the ATS was no longer a statute for foreign-cubed cases, the litigators’ term for suits with a foreign plaintiff, a foreign defendant, and foreign conduct. Filartiga itself had been foreign-cubed in everything except the defendant’s transient presence in Brooklyn, and Kiobel’s logic put even that classic pattern in doubt. The concurring justices, led by Justice Stephen Breyer, would have preserved jurisdiction where the defendant was an American national, the conduct occurred on American soil, or the defendant’s conduct substantially and adversely affected an important American interest. The majority rejected those broader formulations, and the Breyer concurrence became a dissent in all but name.

Kiobel’s practical effect was immediate and dramatic. Dozens of pending ATS cases were dismissed. The plaintiffs’ bar, which had built a practice on foreign-cubed litigation, watched its inventory evaporate. District courts divided over what “touch and concern” required: some held that an American corporation’s mere presence in the United States sufficed, others demanded that the relevant conduct itself occur domestically. The confusion persisted for eight years, until the Court returned to the question in Nestle, and the answer it gave then was narrower than almost anyone had expected.

The decision also revealed the Court’s deeper anxiety, which was not really about statutory interpretation. The presumption against extraterritoriality is a judge-made canon, and applying it to a 1789 statute whose drafters never considered the question was an act of judicial choice, not textual compulsion. The choice reflected a view of the ATS as a diplomatic instrument that had been repurposed into something its makers would not recognize, and a determination to return it to a primarily domestic office. Whether that determination was faithful to the statute or a rewriting of it is the central interpretive dispute of the modern era, and it divides scholars as sharply as it divided the justices.

Touch and Concern: How Courts Apply the Test

How does a court decide whether conduct touches and concerns the United States?

A court looks past the parties’ nationalities and asks where the conduct that violated international law actually happened. If the torture, killing, or other abuse occurred entirely abroad, the claim fails even when the defendant is an American company with headquarters in the United States, because corporate presence alone does not displace the presumption against extraterritoriality.

The full picture requires understanding how the lower courts struggled with the test between 2013 and 2021, because the struggle shows what the Supreme Court later resolved. Some circuits treated “touch and concern” as a sliding scale, weighing the American connections cumulatively: an American defendant plus American decision-making plus American financing might suffice even if the abuse occurred abroad. The Fourth Circuit took this approach in Al Shimari v. CACI, allowing claims by Iraqi detainees against an American military contractor to proceed on the ground that the contractor was American, the contracts were American, and the alleged abuse occurred under American authority. Other circuits demanded that the conduct violating international law itself occur on American soil, a much stricter reading that would have ended nearly every case.

The Supreme Court resolved the split in Nestle USA v. Doe, and the resolution favored the strict reading. But before reaching Nestle, the corporate liability question had to be confronted directly, because Kiobel had been decided on extraterritoriality while leaving open whether corporations could be ATS defendants at all. The Second Circuit had said no. Other circuits had said yes. The resulting division set up Jesner, the case that finally answered the question and, in answering it, revealed how far the Court had traveled from Filartiga.

Can a Corporation Be Sued? Jesner v. Arab Bank

The plaintiffs in Jesner v. Arab Bank were victims of terrorist attacks in Israel, the West Bank, and Gaza, who alleged that Arab Bank, a Jordanian financial institution with a New York branch, had maintained accounts and processed transfers used to finance the attacks. They sued under the ATS, invoking the norm against the financing of terrorism. The Second Circuit dismissed on the ground that corporations cannot be ATS defendants, and the Supreme Court affirmed in 2018 by a five to four vote.

Justice Anthony Kennedy’s majority opinion gave two reasons, and the two reasons point in different directions. The first was historical: Kennedy surveyed the international criminal tribunals, from Nuremberg to the modern ad hoc courts, and found that none had exercised jurisdiction over corporations as entities. The ATS, which enforces the law of nations, could not impose liability on a category of defendant that the law of nations itself did not recognize. The second reason was institutional: imposing corporate liability would have significant diplomatic consequences, requiring courts to hold foreign corporations accountable in ways that could provoke retaliation against American companies abroad, and that choice belonged to Congress, not the judiciary.

The dissent, written by Justice Sotomayor and joined by three colleagues, argued that the majority had the history backwards. The law of nations had never affirmatively excluded corporate liability, the dissenters said, and the question was one of remedy, which domestic law supplies, rather than of the international norm itself. Torture is prohibited whether the torturer is a person or a company, and nothing in the law of nations immunizes an entity that commits it. The dissent also noted the practical consequence: the decision created a perverse incentive, allowing wrongdoers to launder liability through corporate form.

Jesner’s holding was carefully limited to foreign corporations. Kennedy’s opinion repeatedly emphasized that the defendant was a Jordanian bank, and the diplomatic-consequences reasoning applied most forcefully to hauling foreign companies into American courts. The opinion left open whether American corporations could be ATS defendants, and plaintiffs’ lawyers immediately tested the opening. If the defendant was Nestle USA, an American company, Jesner’s reasoning about foreign defendants would not apply, and the case could proceed. That test produced Nestle USA v. Doe, and the result disappointed everyone who had counted on the distinction.

Nestle USA v. Doe: The Domestic Conduct Requirement

The Nestle plaintiffs were former child slaves on cocoa plantations in Ivory Coast, who alleged that Nestle USA and Cargill had aided and abetted child slavery by providing financial and technical support to Ivorian farms while knowing of the labor practices. The companies were American. The plaintiffs argued that Jesner’s foreign-corporation holding did not apply, and that the companies’ American decision-making, the headquarters approvals and supply-chain management in the United States, supplied the domestic conduct that Kiobel’s touch-and-concern test required.

The Supreme Court rejected the argument in 2021, in a fractured decision whose clearest holding came from Justice Clarence Thomas’s plurality. Thomas wrote that the relevant conduct, the alleged aiding and abetting of child slavery, occurred in Ivory Coast, and that general corporate presence and domestic decision-making could not transform foreign abuse into domestic conduct. To hold otherwise, Thomas reasoned, would allow plaintiffs to plead their way around Kiobel by alleging that some corporate planning occurred in an American boardroom, which in a globalized economy would be true of nearly every multinational case. The touch-and-concern test required that the conduct violating international law occur in the United States, not merely that the defendant be American or that business decisions be made here.

The decision’s multiple opinions revealed a Court deeply divided on the ATS’s remaining scope but united on the outcome. Justice Gorsuch, concurring, would have gone further and held that courts lack authority to recognize any new ATS causes of action at all, foreshadowing the position the majority would adopt five years later in Cisco. Justice Sotomayor, concurring in part, agreed that the plaintiffs’ allegations were insufficient but preserved the possibility that domestic conduct could support a claim in a future case. The practical message was unmistakable: after Nestle, an ATS plaintiff needed to show that the violation of international law itself happened on American soil, a requirement that excluded virtually the entire docket the human rights bar had built since Filartiga.

Nestle also illustrated the aiding-and-abetting problem that would dominate the final chapter. The plaintiffs did not allege that Nestle enslaved children. They alleged that Nestle aided and abetted Ivorian farmers who did. That theory, borrowed from criminal law and the Unocal litigation, had carried the corporate cases for two decades without ever receiving the Supreme Court’s endorsement. In Cisco, the Court finally addressed it directly, and the answer closed the last significant opening.

Aiding and Abetting: The Last Great Theory

After Jesner closed the door to foreign corporations and Nestle closed the workaround of domestic decision-making, one theory still carried the remaining corporate cases: aiding and abetting. The idea was straightforward and, in the plaintiffs’ telling, morally obvious. A technology company that sells surveillance equipment knowing it will be used to track dissidents, or a resource company that pays a foreign army knowing the money funds atrocities, has contributed to the abuse. Criminal law has punished accomplices for centuries. International criminal tribunals, from Nuremberg to The Hague, have convicted aiders and abettors. Why should civil liability be different.

The doctrinal problem was that neither the ATS nor the TVPA mentions aiding and abetting. The ATS speaks of torts committed in violation of the law of nations, and the TVPA speaks of an individual who subjects another to torture. Plaintiffs argued that secondary liability was part of the federal common law that Sosa authorized courts to apply, and several circuits agreed, developing tests borrowed from international criminal law. The Second Circuit in the Unocal litigation and later cases asked whether the defendant provided knowing practical assistance that had a substantial effect on the abuse. The standard sounded workable, and it kept the corporate docket alive through the 2010s.

The defense bar attacked the theory on two fronts. First, they argued that the law of nations did not recognize civil aiding-and-abetting liability with the specificity Sosa required, pointing out that the international precedents were criminal, not civil, and that transplanting criminal doctrines into tort law was exactly the kind of judicial invention Sosa warned against. Second, they argued that the TVPA’s text foreclosed the theory for the statutory claims: Congress had written “an individual who subjects” another to torture, language of direct perpetration, and courts should not add words Congress omitted. The two arguments converged in the Cisco litigation, which presented both an ATS aiding-and-abetting claim and a TVPA aiding-and-abetting claim in a single case, giving the Supreme Court the vehicle to resolve the theory’s fate across both statutes at once.

Cisco Systems v. Doe: The Door Closes

The Cisco case began, like so many ATS cases, with victims who had nowhere else to go. The plaintiffs were practitioners of Falun Gong, the spiritual movement banned by the Chinese government, who alleged that they had been subjected to surveillance, detention, and torture by Chinese authorities. They alleged that Cisco Systems, the San Jose networking company, and several of its executives had aided the persecution by designing and selling networking and surveillance infrastructure, known as the Golden Shield project, that Chinese security services used to identify and track practitioners. Cisco denied that its technology had been customized to enable repression, describing its products as standard networking equipment sold worldwide. The plaintiffs sued under both the ATS and the TVPA, invoking aiding-and-abetting liability under each. The Ninth Circuit allowed the claims to proceed. The Supreme Court reversed.

The decision, Cisco Systems, Inc. v. Doe, No. 24-856, was six to three. Justice Amy Coney Barrett wrote for the majority, joined by Chief Justice Roberts and Justices Thomas, Alito, Gorsuch, and Kavanaugh. The opinion made two holdings, each aimed at one of the two statutes. On the ATS, Barrett held that federal courts may not create new causes of action for violations of international norms, closing the door that Sosa had left ajar. Sosa had permitted courts to recognize claims based on norms as definite and universal as the eighteenth-century paradigms; Barrett’s majority concluded that the recognition power itself was the problem, because creating a cause of action is, in her phrase, an extraordinary act that places great stress on the separation of powers, a power that belongs to Congress. On the TVPA, the Court held that the statute, which contains an express cause of action, does not provide for aiding-and-abetting liability. Congress wrote liability for the individual who subjects another to torture, and courts may not expand the text to reach those who assisted.

The opinion states the first holding in a single sentence: the Court closes the door that Sosa cracked and holds that courts may not create new causes of action for violations of international norms. The sentence is doing constitutional work disguised as statutory interpretation. Its premise is that every new ATS claim is an exercise of legislative power, defining prohibited conduct and assigning liability, and that the judiciary may not exercise legislative power in the service of foreign policy no matter how worthy the cause. The three dissenters, whose opinions were not detailed in the initial reporting, evidently took the opposite view: that Congress in 1789 had authorized exactly this kind of interstitial lawmaking and that the majority was rewriting the statute’s history to serve a cramped vision of judicial power.

The decision’s practical effect was to end the ATS as a vehicle for new claims. Existing causes of action recognized before Cisco, the narrow set that had survived Sosa and Kiobel, were not directly disturbed, but the opinion’s logic leaves them resting on a foundation the majority has repudiated. And the TVPA holding ended the last corporate theory, since aiding-and-abetting had been the mechanism for reaching companies that did not themselves commit torture. After Cisco, a plaintiff must show that the defendant personally subjected the victim to torture or extrajudicial killing, a requirement that excludes the supply chain, the financing channel, and the technology sale. The human rights bar’s four-decade project of building transnational corporate accountability through American courts reached its terminus.

The table below compresses the arc this article has traced. Each row is a turning point, each column says what changed, and the whole fits on a single glance for the reader who wants the doctrine’s shape without the narrative.

Turning point	Year	What it changed
Filartiga v. Pena-Irala	1980	Revived the statute; official torture held to violate the law of nations
Torture Victim Protection Act	1991	Congress created an express remedy for torture and extrajudicial killing
Sosa v. Alvarez-Machain	2004	Courts may recognize only claims based on specific universal obligatory norms
Kiobel v. Royal Dutch Petroleum	2013	Presumption against extraterritoriality; claims must touch and concern the United States
Jesner v. Arab Bank	2018	Foreign corporations cannot be sued under the statute
Nestle USA v. Doe	2021	Domestic corporate presence is not enough; the violating conduct must be domestic
Cisco Systems v. Doe	2026	Courts may not create new ATS causes of action; TVPA has no aiding-abetting liability

The Strongest Case for the Open Door

An article that never argues against its own thesis is marketing, and the thesis of this article, that only Congress may open the courthouse to the world’s victims, has always had formidable opponents. Their case deserves its best form, because it is the case that carried the statute through its heroic age and that still moves judges, scholars, and legislators.

The advocates’ argument begins with the forum problem. Torture, genocide, and crimes against humanity are typically committed by governments or with government complicity, in countries whose courts are captured, collapsed, or complicit. Harold Koh, the Yale professor who served as the State Department’s legal adviser, has argued for decades that transnational litigation supplies the forum that the international system lacks. There is no world court for civil damages. The International Criminal Court prosecutes individuals criminally and slowly, and its jurisdiction is limited. Regional human rights courts issue judgments that states often ignore. For a victim in a country whose judiciary answers to the torturer’s superiors, an American courtroom may be the only room in the world where the claim can be heard. Closing the ATS door does not return these victims to an adequate alternative forum. It returns them to no forum at all.

The second strand of the argument is that the separation-of-powers objection proves too much. American courts routinely decide cases with foreign policy implications. They interpret treaties, apply foreign law, hear suits against foreign sovereigns under the Foreign Sovereign Immunities Act’s terrorism exception, and adjudicate the rights of aliens in immigration and extradition proceedings. The political question doctrine and the act of state doctrine already give judges tools to avoid cases that would truly embarrass the executive. Beth Stephens, the human rights litigator and scholar who argued some of the landmark ATS cases, has contended that the judiciary’s competence to handle sensitive transnational cases is demonstrated by decades of practice, and that the skeptics’ parade of horribles, courts conducting foreign policy, American companies fleeing difficult markets, diplomatic retaliation, never materialized during the statute’s most expansive years.

The third strand is textual and historical. The advocates read the First Congress as having authorized a dynamic statute, pointing to Kaufman’s reasoning in Filartiga and to Souter’s acknowledgment in Sosa that Congress expected the law of nations to evolve. On this reading, the Cisco majority did not interpret the statute but rewrote it, replacing the founders’ design with a modern preference for judicial restraint. William Casto’s historical work, though often cited by both sides, contains material for the advocates: the First Congress cared about providing a federal forum for law-of-nations violations, and the specific content of those violations was always going to change with the law itself.

The fourth strand is moral and is stated most plainly by the victims’ lawyers: a legal system that can hear a slip-and-fall case but not a torture case has its priorities backwards. The ATS at its height expressed an American commitment, however imperfect, to the proposition that some wrongs are so grave that no border shields them. That commitment had expressive value beyond the judgments collected, signaling to perpetrators that travel to the United States carried risk and to victims that their suffering was legible to law. The narrowing of the statute, in this telling, is not doctrinal hygiene but moral retreat, a decision that the comfort of American courts matters more than the claims of the world’s brutalized.

The Defenders’ Answer and Where the Thesis Holds

The defenders of the narrowing, the position this article’s thesis supports, have answers to each strand, and the answers are stronger than the advocates usually admit. To the forum problem, the defenders respond that the absence of a perfect forum does not authorize American judges to invent one. The Constitution assigns foreign affairs to the political branches precisely because those branches can weigh competing interests, diplomatic relationships, economic consequences, and strategic costs, that courts cannot measure and should not balance. Curtis Bradley’s scholarship makes this point systematically: transnational litigation is foreign policy by other means, and foreign policy without democratic accountability is not justice but judicial adventurism. If the world needs a civil forum for human rights claims, the remedy is a treaty negotiated by the executive and ratified by the Senate, not a judge’s improvisation.

To the competence argument, the defenders respond that the political question doctrine is a shield, not a license. The fact that courts can avoid the worst cases does not mean they should hear the rest, because the very filing of an ATS suit against a foreign corporation or official creates diplomatic friction regardless of the outcome. The State Department repeatedly told the courts exactly this, filing statements of interest in ATS cases warning that the litigation complicated American foreign relations. In Kiobel, the governments of the United Kingdom and the Netherlands filed briefs opposing ATS jurisdiction over their corporations, a remarkable spectacle: America’s closest allies asking its highest court to restrain its lower courts. A doctrine that alarms allies is not obviously a doctrine that serves American interests.

To the historical argument, the defenders respond with the history itself. The First Congress that passed the ATS also passed the Judiciary Act’s other provisions, and nothing in the debates suggests an intent to create a global tort forum. The dynamic reading that Kaufman adopted in Filartiga was, on this view, an invention of the late twentieth century, projected backward onto founders who would not have recognized it. Souter’s compromise in Sosa, generous as it was to the advocates, already contained the concession that courts must exercise “restraint” and consider “practical consequences,” language that the later majorities simply took seriously.

To the moral argument, the defenders offer the hardest answer: that law is not the only instrument of morality, and that courts are not the only venue for justice. Diplomacy, sanctions, foreign aid conditionality, international criminal prosecution, and domestic reform movements all address human rights abuses, and each operates with tools courts lack. The moral urgency of torture does not confer jurisdiction, because jurisdiction is a question of power, not of desert. A court that hears every worthy claim becomes a legislature, and a legislature of unelected judges is not a moral achievement.

Where does this leave the thesis? The defenders’ case is strongest on the constitutional ground and weakest on the practical one. It is true that creating causes of action is legislative power, and it is true that the ATS’s text does not confer it. But it is also true that Congress has shown no inclination to build the forum the advocates want, which means the constitutional argument functions, in practice, as an argument for doing nothing. The honest verdict is the one this article’s thesis states: the narrowing is constitutionally defensible and humanly costly, and both of those things can be true at once. The law chose the separation of powers over the victims’ forum. That is a choice with reasons, not a mistake with excuses.

The Procedural Gauntlet: Jurisdiction, Forum, Politics

Even before the Supreme Court narrowed the substantive law, ATS plaintiffs ran a procedural gauntlet that defeated most cases, and the gauntlet survives for the claims that remain. Understanding it is essential, because the popular image of human rights litigation, a victim telling a story to a jury, omits the years of motion practice that determine whether any story gets told.

Personal jurisdiction comes first. A court must have power over the defendant, which traditionally requires that the defendant be served with process inside the forum or have sufficient contacts with it. Filartiga’s defendant was served in Brooklyn while awaiting deportation, the classic transient presence. Modern corporate defendants are served through their American subsidiaries or agents, which generates fights over whether a parent company is subject to jurisdiction based on a subsidiary’s contacts. The Supreme Court’s decisions in Goodyear (2011) and Daimler (2014) narrowed general jurisdiction over corporations to their states of incorporation and principal places of business, which removed one more path for reaching foreign companies.

Forum non conveniens comes second. Even when jurisdiction exists, a court may dismiss if a foreign forum is adequate and more convenient, weighing the private interests of the parties and the public interest in the dispute. Defendants routinely argued that the courts of the country where the abuse occurred were the natural forum. Plaintiffs responded that those courts were precisely the problem, and courts developed a sliding scale: the more the foreign judiciary was shown to be compromised, the less weight its adequacy received. The doctrine produced extensive satellite litigation over the quality of foreign justice systems, with expert declarations about judicial independence in countries from Burma to Colombia.

The political question doctrine and the act of state doctrine come third. The political question doctrine holds that some disputes are committed to the elected branches and are therefore nonjusticiable. The act of state doctrine holds that American courts will not judge the validity of a foreign sovereign’s acts within its own territory. Defendants invoked both to argue that ATS suits second-guessed American foreign policy or condemned allied governments. Courts applied them sparingly but not never: cases directly challenging active American military or intelligence operations were the most likely to be dismissed, while cases against deposed regimes or private actors usually survived.

The Foreign Sovereign Immunities Act adds a fourth barrier for suits against states themselves. The FSIA, enacted in 1976 and codified at 28 U.S.C. sections 1602 through 1611, provides that foreign states are immune from American jurisdiction subject to enumerated exceptions, including the commercial activity exception and the terrorism exception for designated state sponsors. The ATS does not override FSIA immunity, as the Supreme Court confirmed in Argentine Republic v. Amerada Hess (1989). A victim seeking to sue the torturing state itself, rather than the individual torturer, must fit within an FSIA exception, which is a separate and demanding inquiry.

The cumulative effect of the gauntlet was that even at the statute’s height, only a fraction of filed cases reached discovery, and only a fraction of those reached trial. The procedure was the substance: each doctrine expressed a judgment about the proper limits of American judicial power, and together they ensured that transnational litigation remained exceptional rather than routine.

Remedies: What Victory Looks Like

Why do most of these cases end in default judgments?

Most of these cases end in default judgments because the defendants, foreign officials and entities with no reason to respect an American court’s authority, refuse to appear, so the court enters judgment on the plaintiff’s uncontested allegations without a trial, producing large awards that are rarely collectible.

The remedial picture is the statute’s great paradox: sweeping legal victories paired with empty practical results. The Filartiga judgment exceeded ten million dollars and was never paid. The Marcos judgment approached two billion dollars and produced decades of asset-chasing that recovered a fraction. The pattern repeated across the docket: default judgments against absent defendants, dramatic numbers, minimal collection. Scholars of the field divide over whether this matters. The expressive theorists argue that the judgment itself is the remedy, a formal declaration by a court of law that the abuse happened and was wrongful, with value independent of payment. The compensation theorists respond that unenforceable judgments risk turning human rights law into theater, teaching perpetrators that American courts bark but do not bite.

Damages in the cases that did pay came from three sources. First, defendants with American assets: the Marcos estate’s Hawaii proceedings recovered funds because the family had parked wealth in the United States. Second, settlements: Unocal settled on the eve of trial, and several corporate defendants settled to avoid discovery into their foreign operations, with the settlement amounts often sealed. Third, the rare contested trial: the 1994 Marcos jury verdict was contested and resulted in real, if partial, recovery through years of enforcement proceedings.

Beyond money, courts awarded declaratory relief, formal statements that the defendant’s conduct violated international law, and in a few cases injunctive relief. Punitive damages were available and frequently awarded in large amounts, reflecting juries’ moral condemnation. Attorney’s fees, however, were generally not recoverable, which meant the plaintiffs’ bar worked on contingency or through nonprofit funding, a financial structure that shaped which cases were brought. A case needed a solvent defendant or a wealthy backer, which is why the docket tilted toward corporate defendants and deposed dictators with traceable assets, and away from the judgment-proof warlords whose victims arguably needed the forum most.

The remedial limits also explain why the corporate cases mattered so disproportionately. A judgment against a multinational company is collectible in a way that a judgment against a retired torturer is not. When Jesner, Nestle, and Cisco closed the corporate theories, they did not merely narrow the doctrine. They removed the defendants against whom judgments could actually be enforced, leaving a remedy that works, in the main, only against individuals foolish enough to travel to the United States with assets to seize.

Beyond America: How Other Countries Handle These Claims

The United States was never the only country to experiment with transnational human rights litigation, and the comparative picture puts the ATS story in perspective. The experiments elsewhere took different institutional forms, and their fates rhyme with the American one.

Belgium tried the most ambitious version. Its 1993 universal jurisdiction law allowed Belgian courts to prosecute grave breaches of international humanitarian law regardless of where they occurred or the nationality of the parties, and private victims could initiate cases. The law produced complaints against sitting heads of state, including Ariel Sharon of Israel and George H. W. Bush of the United States, which generated intense diplomatic pressure. In 2003, Belgium amended the law to require a Belgian connection and transferred prosecutorial discretion to the federal prosecutor, effectively ending the experiment. The Belgian story is the ATS story in fast-forward: an expansive opening, diplomatic backlash, legislative retrenchment.

Spain’s Audiencia Nacional pursued universal jurisdiction investigations into abuses in Guatemala, Tibet, and elsewhere, led by the crusading judge Baltasar Garzon. The Spanish Constitutional Court initially supported the broad reading, but legislative reforms in 2009 and 2014 narrowed the jurisdiction to cases with Spanish victims or defendants present in Spain. The Netherlands has convicted individuals for war crimes committed abroad, including the 2023 conviction of a former Syrian militia member, but its cases are criminal prosecutions brought by the state, not civil suits brought by victims. France’s specialized war crimes unit has pursued cases against alleged Syrian torturers under its universal jurisdiction provisions, again through criminal prosecution.

The United Kingdom, the other great common-law jurisdiction, never developed an ATS equivalent. English courts hear transnational tort claims under ordinary conflict-of-laws principles, and the landmark Vedanta (2019) and Okpabi (2021) decisions of the UK Supreme Court allowed claims by foreign plaintiffs against British parent companies for overseas harms, but on negligence grounds under domestic tort law, not on international law grounds. The English approach achieves some of what the ATS attempted, corporate accountability for foreign abuses, without invoking the law of nations at all.

The comparative lesson is that every jurisdiction that opened a broad transnational forum eventually narrowed it under diplomatic and practical pressure, and that the narrowing took similar forms: connection requirements, prosecutorial discretion, corporate limitations. The American version was distinctive only in its mechanism, a civil statute wielded by private plaintiffs, and in its duration, four decades of expansion before the contraction. The world’s appetite for a global courthouse appears to be limited everywhere, not just in Washington.

Where Victims Can Still Go

The closing of the ATS door raises the practical question that matters most to the reader: what remains. The answer is a patchwork, less than the advocates want and more than the skeptics admit, and its pieces deserve an honest inventory.

The TVPA survives for its core cases. A victim of torture or extrajudicial killing committed under color of foreign law may still sue the individual perpetrator in American courts, provided the perpetrator can be served and the ten-year limitations period is met. After Cisco, the suit must target the direct perpetrator, not the aider or abettor, which focuses the litigation on officials and commanders rather than financiers and suppliers. The TVPA’s exhaustion requirement and its limitation to natural persons remain, but the statute is, as Justice Sotomayor’s Mohamad opinion confirmed, a genuine congressional remedy, and no court can narrow what Congress wrote.

Criminal prosecution remains available where the political will exists. The United States prosecutes torture committed abroad by American nationals under 18 U.S.C. section 2340A, the federal torture statute, and has used it in cases arising from the war on terror. Foreign officials who travel to the United States have been prosecuted for immigration fraud based on concealed human rights abuses, a creative workaround that has put several perpetrators in American prisons. Internationally, the International Criminal Court continues its work, however slowly and selectively, and national courts exercising universal jurisdiction continue to bring criminal cases.

Regional human rights courts offer another path. The European Court of Human Rights, the Inter-American Court of Human Rights, and the African Court on Human and Peoples’ Rights hear cases against states, not individuals, and award remedies that states sometimes implement. Their judgments lack the ATS’s direct enforceability against perpetrators, but they produce authoritative findings that shape diplomacy and domestic politics.

Civil litigation in other countries’ courts is growing as the American forum contracts. The English negligence cases, Dutch and French criminal-civil hybrids, and Canadian transnational tort litigation all offer partial substitutes. None replicates the ATS’s combination of American discovery, jury trials, and punitive damages, which is why the plaintiffs’ bar prized it, but the direction of travel is toward a decentralized system of national forums rather than a single global courthouse.

Finally, nonjudicial mechanisms matter more than lawyers admit. Sanctions programs, including the Global Magnitsky sanctions administered by the United States Treasury, freeze the assets and restrict the travel of designated human rights abusers, achieving concrete consequences without a courtroom. Truth commissions, lustration programs, and domestic prosecutions in transitional societies have delivered forms of accountability that civil damages cannot. The debate over accountability after atrocity is wider than the debate over any single statute, and the statute’s closing does not end it. The worldwide reckoning with state violence that followed a consulate killing showed how diplomacy, journalism, sanctions, and courts combine when no single forum suffices, and that combination, messy as it is, is the realistic future of transnational accountability.

The Hard Cases at the Edges

Every doctrine has edge cases that test its logic, and the ATS produced several that remain instructive even after the statute’s effective closing. They are worth examining because they show where the law’s categories strained, and the strain reveals the categories’ shape.

The first edge case is the American defendant who commits abuse abroad. The ATS requires an alien plaintiff but says nothing about the defendant’s nationality, so an American official who tortures a foreigner overseas could in principle be an ATS defendant. In practice, such cases were rare, because American officials acting abroad usually acted under government authority, which raised immunity and political question defenses, and because plaintiffs preferred the TVPA or constitutional claims. The edge case matters theoretically: it shows that the statute’s alien-plaintiff requirement was about giving foreigners a forum, not about regulating Americans, a one-way design that the modern docket’s focus on corporate defendants partially obscured.

The second edge case is the norm that almost qualified. Litigants proposed dozens of candidates under Sosa: forced labor, environmental destruction, cruel and degrading treatment, prolonged arbitrary detention, corporate complicity in war crimes. Courts rejected nearly all of them, but the rejections were often fact-specific, leaving open whether a better-defined version might succeed. The prohibition of child slavery, for example, is arguably as specific and universal as the prohibition of torture, and the Nestle plaintiffs framed their claim accordingly. The Court rejected the claim on extraterritoriality grounds without reaching the Sosa question, so the norm’s status remains technically open. After Cisco, the openness is academic for new ATS claims, but it matters for understanding what Sosa left undecided.

The third edge case is the treaty prong, the statute’s forgotten half. The ATS covers torts committed in violation of a treaty of the United States, but almost no litigation developed under this prong, because few treaties create privately enforceable rights and courts construe the self-execution question strictly. The prong’s dormancy is a standing invitation to scholars: a future treaty with an express private right of action could revive it, though no such treaty is on the horizon.

The fourth edge case is the existing recognized claim. Cisco held that courts may not create new causes of action, but it did not purport to overrule the claims recognized before, the narrow set that survived Sosa and Kiobel. A future plaintiff alleging official torture with strong domestic connections might still invoke those precedents. Whether lower courts will treat the surviving claims as viable or as dead letters resting on repudiated reasoning is the most genuinely open question in the field, and the answer will emerge only as cases are filed and decided.

What Is Settled and What Remains Open

The doctrine’s current state can be stated with unusual confidence for a field this contested, because the Supreme Court has addressed the field so comprehensively. What follows is the map as it stands, with the settled and the open clearly marked.

Settled: the ATS grants jurisdiction only and creates no cause of action. Settled: federal courts may not recognize new ATS causes of action for violations of international norms. Settled: the presumption against extraterritoriality applies, and claims must be based on conduct that occurred in the United States. Settled: foreign corporations cannot be ATS defendants. Settled: general corporate presence and domestic decision-making do not satisfy the domestic-conduct requirement. Settled: the TVPA provides no aiding-and-abetting liability and runs only against natural persons. Settled: the TVPA requires exhaustion of foreign remedies and carries a ten-year limitations period.

Open: whether the causes of action recognized before Cisco retain vitality. Open: whether any ATS claim with genuinely domestic conduct, abuse committed on American soil violating the law of nations, could still proceed, a scenario the cases have never squarely presented. Open: how courts will treat the treaty prong if a suitable treaty appears. Open: whether Congress will revisit the field, either by amending the ATS to create express causes of action or by expanding the TVPA, a legislative step that human rights advocates continue to urge and that has never advanced beyond proposal.

The table introduced earlier compresses this history into its turning points, and it is worth a second glance now that the whole argument is on the table. Read the “what it changed” column from top to bottom: revival, legislation, discipline, territorial limit, defendant limit, conduct limit, closure. Each row narrowed the opening the previous row had left. The pattern is not accidental. It is the shape of a judiciary progressively deciding that the statute’s promise exceeded its constitutional warrant, and acting, decision by decision, to bring the promise back within the warrant.

For readers who want to keep the case sequence at hand while studying the doctrine’s development, a companion reference tool is available alongside this article.

Why the Argument Will Outlive the Statute’s Answers

The Alien Tort Statute began as a diplomatic courtesy and became, for four decades, the most ambitious instrument of transnational justice ever placed in private hands. It ends, for practical purposes, as a cautionary tale about the limits of judicial power: a statute that promised the world’s victims an American courtroom and delivered, instead, a master class in why courtrooms cannot do the work of foreign policy. The thesis with which this article began has survived its own complication. American courts may hear the world’s worst cases only when Congress clearly says so, because inventing remedies for foreign wrongs is governing, and governing belongs to the branches the people elect.

That verdict should not be mistaken for indifference. The abuses the statute addressed are real, the victims’ need for a forum is real, and the moral cost of the narrowing is real. A legal order that closes its courts to torture victims has made a choice, and the choice deserves to be named rather than disguised as technicality. But the choice also has reasons that survive scrutiny. A judiciary that creates causes of action from the evolving customs of the international system exercises a power no constitution gave it, over disputes no election authorized it to resolve, with consequences no judge can fully foresee. The separation of powers is not a technicality either. It is the mechanism by which a republic keeps its promises to its own citizens while deciding what it owes to the world’s.

The argument will outlive the statute because the underlying tension is permanent. Every generation rediscovers the impulse to use law against atrocity and rediscovers, in turn, the limits of what law can do across borders. New instruments will be proposed: broader statutes, new treaties, expanded criminal jurisdiction, stronger sanctions. Some will succeed where the ATS failed. All will face the same questions the ATS faced, about who decides, by what authority, and at what cost to the constitutional order that makes the decision possible. The statute’s two centuries of life, from a Philadelphia street assault in 1784 to a Supreme Court opinion in 2026, are the fullest record we have of how those questions get answered in practice. The record is closed. The questions are not. A companion research aid is linked here for readers tracing how those questions recur across the archive.

Frequently Asked Questions

Q: What is the Alien Tort Statute in plain language?

The Alien Tort Statute is a 1789 federal law, thirty-three words long, that gives American federal courts the power to hear civil lawsuits brought by noncitizens for wrongs committed in violation of international law or an American treaty. In plain terms, it opens an American courthouse door to foreigners who suffered grave abuses abroad. The critical subtlety is that the statute grants only the power to hear the case, not the underlying right to win it. For two centuries courts have argued over what claims can walk through that door. The modern answer, settled across six Supreme Court decisions, is: very few. The statute still exists, but judges may no longer invent new claims under it, and surviving claims must involve conduct inside the United States.

Q: Who can file a lawsuit under the Alien Tort Statute?

Only a noncitizen of the United States, called an alien in the statute’s eighteenth-century vocabulary, may file an Alien Tort Statute claim. An American citizen tortured abroad cannot use this law and must look elsewhere, such as the Torture Victim Protection Act, which has no citizenship requirement for plaintiffs. The plaintiff must allege a tort, meaning a civil wrong like assault, battery, or false imprisonment, rather than a breach of contract or some other kind of dispute. The tort must violate the law of nations, the body of customary international law, or a treaty of the United States. In practice, the plaintiffs were refugees, asylees, and exiles living in America who had suffered abuse in their home countries and discovered that their abuser was reachable by an American court.

Q: Who can be sued under the Alien Tort Statute?

The defendants have included foreign officials, deposed dictators, private militia leaders, and multinational corporations, but each category has been narrowed by the Supreme Court. Individual perpetrators, including officials acting under color of foreign law, were the classic defendants from the Filartiga era onward. Foreign corporations were held to be immune from ATS suits in Jesner v. Arab Bank in 2018. American corporations remained theoretically reachable until Nestle USA v. Doe in 2021 required the violating conduct itself to occur domestically, and Cisco Systems v. Doe in 2026 barred courts from creating new causes of action at all. Foreign governments themselves generally cannot be sued under the statute because the Foreign Sovereign Immunities Act of 1976 grants them immunity subject to narrow exceptions.

Q: What kinds of abuses count as violations of the law of nations?

Courts have recognized only a short list: official torture, extrajudicial killing, genocide, war crimes, and crimes against humanity, with piracy and offenses against ambassadors surviving from the eighteenth-century originals. The governing standard comes from Sosa v. Alvarez-Machain in 2004, which requires the norm to be specific, universal, and obligatory, defined with a clarity comparable to the historic paradigms. Torture qualifies because the Convention Against Torture defines it precisely and more than one hundred seventy states have condemned it. Broader or vaguer candidates, such as arbitrary detention, environmental harm, or forced labor framed as cruel treatment, have generally failed the test. The list has not grown in decades, and after the 2026 Cisco decision courts may not add to it.

Q: Why did the Alien Tort Statute go unused for almost 200 years?

The statute slept from 1789 to 1980 because the legal world it needed did not yet exist. Nineteenth-century international law governed relations between states, not the rights of individuals, so the idea that a government could violate international law by torturing its own citizens had no legal footing. American courts had no occasion to stretch the law, since its original purposes, protecting ambassadors and punishing pirates, were handled through criminal and admiralty jurisdiction. The human rights revolution after the Second World War changed the picture: the Universal Declaration of Human Rights in 1948, the covenants of the 1960s, and the Convention Against Torture in 1984 crystallized prohibitions that courts could apply. A generation of civil rights lawyers then supplied the litigation strategy, and the Filartiga family’s case supplied the facts.

Q: What happened in Filartiga v. Pena-Irala?

In 1976, Paraguayan police under the Stroessner dictatorship tortured to death the teenage son of dissident physician Joel Filartiga. The family later found the alleged torturer, police inspector Americo Norberto Pena-Irala, living in Brooklyn, New York, and sued him in federal court under the Alien Tort Statute. The trial court dismissed, but the Second Circuit reversed in 1980, with Judge Irving Kaufman holding that official torture violated the law of nations and that the statute must be read against modern international law rather than the law of 1789. On remand the court entered a default judgment exceeding ten million dollars, which was never collected. The case mattered not for the money but for the precedent: it revived a dormant statute and launched four decades of transnational human rights litigation in American courts.

Q: What is the Sosa test for Alien Tort Statute claims?

The Sosa test comes from Sosa v. Alvarez-Machain, decided by the Supreme Court in 2004, and it controls which claims courts may recognize under the statute. A modern claim is actionable only if it rests on a norm of international character accepted by the civilized world and defined with a specificity comparable to the eighteenth-century paradigms of piracy, offenses against ambassadors, and violations of safe conduct. In practice the test works as a funnel: torture passes because the Convention Against Torture defines it with precision and the prohibition is universal, while vaguer claims like arbitrary detention fail for lack of definitional sharpness. Courts must also weigh the practical consequences of recognizing a claim, including its effect on foreign relations. Few proposed norms have survived this scrutiny, and the 2026 Cisco decision closed the recognition power entirely.

Q: Can corporations be sued under the Alien Tort Statute?

As a practical matter, no longer. The Supreme Court closed the corporate theories in stages. Jesner v. Arab Bank in 2018 held that foreign corporations cannot be ATS defendants, reasoning that the law of nations as enforced by international tribunals had never recognized corporate liability and that the diplomatic consequences of such suits belonged to Congress to weigh. Nestle USA v. Doe in 2021 addressed American corporations and held that the conduct violating international law must itself occur in the United States, so headquarters planning and financing could not support claims about overseas abuse. Cisco Systems v. Doe in 2026 then barred aiding-and-abetting liability and new causes of action altogether. A corporation that directly commits torture on American soil in violation of international law remains theoretically reachable, but that scenario has never produced a reported case.

Q: What did Kiobel v. Royal Dutch Petroleum decide?

Kiobel v. Royal Dutch Petroleum, decided in 2013, held that the presumption against extraterritoriality applies to the Alien Tort Statute, meaning the law is presumed to govern only inside the United States unless Congress clearly says otherwise. Nigerian plaintiffs had sued Dutch and British oil companies over alleged complicity in abuses by the Nigerian military, with almost no American connection. Chief Justice Roberts, writing for the majority, held that such foreign-cubed claims, foreign plaintiff, foreign defendant, foreign conduct, must be dismissed unless they touch and concern American territory with sufficient force to displace the presumption. The decision ended the era in which American courts served as a global forum for foreign disputes and forced plaintiffs to show genuine domestic connections, a requirement the Court tightened further in later cases.

Q: What is the Torture Victim Protection Act?

The Torture Victim Protection Act of 1991 is the one occasion on which Congress built its own remedy for overseas abuse instead of leaving the work to judges. Signed in 1992, it provides that any individual who, under actual or apparent authority of a foreign nation, subjects another person to torture or extrajudicial killing shall be liable for damages in an American court. Unlike the Alien Tort Statute, it creates an express cause of action, so courts enforcing it carry out Congress’s command rather than inventing their own. It covers only two abuses, requires plaintiffs to exhaust adequate remedies where the abuse occurred, runs only against natural persons rather than organizations, and carries a ten-year limitations period. It remains the most viable statutory path for victims after the ATS narrowing.

Q: How is the Torture Victim Protection Act different from the Alien Tort Statute?

The two laws differ in source, scope, plaintiffs, defendants, and procedure. The ATS is a 1789 jurisdictional grant that covers any tort violating the law of nations but creates no cause of action of its own, forcing courts to recognize claims under strict tests. The TVPA is a 1991 statute that expressly creates liability but only for torture and extrajudicial killing. The ATS admits only alien plaintiffs, while the TVPA admits anyone, including Americans. The ATS historically reached corporations until the Supreme Court closed that path, while the TVPA has always run only against natural persons. The TVPA requires exhaustion of foreign remedies and has a ten-year limitations period, neither of which appears in the ATS text. Lawyers once pleaded both together; after the recent decisions, the TVPA’s express remedy is the sturdier of the two.

Q: What does aiding and abetting mean in human rights lawsuits?

Aiding and abetting is secondary liability: holding a party responsible not for committing abuse directly but for knowingly helping the abuser. In the ATS context, plaintiffs used the theory against companies that sold equipment, provided financing, or shared logistics with abusive regimes, arguing that knowing practical assistance with substantial effect on the abuse should create civil liability. The theory borrowed from criminal law and international tribunals, which have long punished accomplices. It carried major cases, including the Unocal pipeline litigation and the Cisco Systems case, for two decades without definitive Supreme Court approval. In Cisco Systems v. Doe in 2026, the Court rejected the theory for both statutes: the ATS, because courts may not create new causes of action, and the TVPA, because Congress wrote liability for those who directly subject victims to torture, not for those who assist.

Q: What did the Supreme Court decide in Cisco Systems v. Doe?

In Cisco Systems v. Doe, decided in 2026 by six votes to three, the Supreme Court closed the last significant openings in transnational human rights litigation. Justice Barrett’s majority opinion held, first, that federal courts may not create new causes of action under the Alien Tort Statute for violations of international norms, ending the recognition power the Court had preserved in Sosa v. Alvarez-Machain. Second, it held that the Torture Victim Protection Act does not provide aiding-and-abetting liability, since Congress imposed liability on the individual who subjects another to torture, not on assistants. The case arose from claims by Falun Gong practitioners that the technology company had aided Chinese persecution through surveillance infrastructure, allegations the company denied. The Ninth Circuit had allowed the claims; the Supreme Court reversed, leaving direct perpetrators as the only reachable defendants.

Q: Can a US court hear a case about torture that happened in another country?

Sometimes, but the openings are narrow and shrinking. The Torture Victim Protection Act allows suits over torture committed abroad under color of foreign law, provided the plaintiff exhausts adequate foreign remedies, sues within ten years, and names the individual perpetrator rather than a company or assistant. The Alien Tort Statute once offered a broader path, but the Supreme Court’s decisions from Kiobel through Cisco require the violating conduct to have occurred in the United States and bar courts from recognizing new claims. Criminal prosecution remains possible in some circumstances, including under the federal torture statute for American nationals and immigration-fraud prosecutions of perpetrators who enter the country. Foreign courts exercising universal jurisdiction and regional human rights courts offer additional, partial alternatives. No single forum replaces what the ATS once promised.

Q: What is the presumption against extraterritoriality?

The presumption against extraterritoriality is a rule of statutory interpretation holding that federal laws are presumed to apply only within the United States unless Congress gives a clear indication that they reach conduct abroad. The rule reflects the principle that American law governs domestically and does not rule the world, along with the practical concern that applying American law to foreign conduct creates friction with other sovereigns. The Supreme Court applied the presumption to the Alien Tort Statute in Kiobel v. Royal Dutch Petroleum in 2013, reasoning that the 1789 text contains no clear statement of overseas reach. The practical result was the touch-and-concern test: claims must be tied to American territory with enough force to displace the presumption, a standard later decisions interpreted to require that the violating conduct itself occur domestically.

Q: What happens if the defendant is a foreign government?

A foreign government sued in an American court is generally protected by the Foreign Sovereign Immunities Act of 1976, which grants immunity subject to specific exceptions. The Alien Tort Statute does not override that immunity, as the Supreme Court confirmed in Argentine Republic v. Amerada Hess in 1989. A victim seeking to sue the torturing state itself must fit within an FSIA exception, such as the commercial activity exception or the terrorism exception that applies to designated state sponsors of terrorism. Individual officials of the foreign state stand on different footing: they may be sued under the TVPA or, historically, the ATS for acts like torture, though immunity doctrines and the narrowing decisions limit those paths as well. The distinction between the state and its officials is one of the most litigated boundaries in the field.

Q: How long do plaintiffs have to file these cases?

The Torture Victim Protection Act sets a ten-year limitations period, with equitable tolling, meaning pause of the clock, for extraordinary circumstances such as the plaintiff’s inability to access courts while a repressive regime remained in power. Courts borrowed the same ten-year period for Alien Tort Statute claims, since the ATS contains no limitations provision of its own. The generous tolling rules recognized the reality that victims of state torture often cannot sue until years after the abuse, when they have escaped the country or the regime has fallen. Cases filed decades after the events, including Holocaust-era and Latin American dirty-war litigation, proceeded under these tolling doctrines. Plaintiffs who delay without justification, however, still lose their claims, and defendants routinely challenge timeliness as a first line of defense.

Q: What compensation can victims win in these cases?

Successful plaintiffs have won compensatory damages for medical costs, lost earnings, and pain and suffering, plus punitive damages reflecting moral condemnation, with awards ranging from millions to the nearly two-billion-dollar Marcos verdict of 1994. The difficulty has always been collection rather than winning. Most defendants were foreign officials who ignored the proceedings, producing default judgments that were never paid. Real recovery came from defendants with reachable assets, such as the Marcos estate’s American holdings, or from settlements, as when Unocal settled on the eve of trial. Courts also granted declaratory relief, formal judicial findings that the abuse violated international law, which victims and advocates valued independently of money. The narrowing decisions removed the corporate defendants against whom judgments were actually collectible, which is one reason the doctrine’s practical value declined faster than its symbolic value.

Q: Why do critics say the statute interferes with foreign policy?

Critics argue that letting judges hear claims about foreign abuses transfers foreign policy from elected officials to the judiciary. Every ATS suit against a foreign official or corporation creates diplomatic friction regardless of its outcome, forcing the State Department to manage fallout from litigation it did not choose. Allied governments, including the United Kingdom and the Netherlands in the Kiobel case, filed briefs asking American courts to restrain themselves. Scholars like Curtis Bradley contend that customary international law is made by foreign governments and scholars rather than by any legislature accountable to American voters, so enforcing it as federal law lets the world legislate for the United States. The business community added that the suits punished American companies for operating in difficult countries. Defenders reply that courts already handle sensitive transnational cases and that the predicted harms never fully materialized, but the constitutional argument carried the Supreme Court.

Q: Where else in the world can victims bring human rights cases?

Several alternatives exist, each partial. Criminal prosecution under universal jurisdiction continues in countries like France, Germany, and the Netherlands, where specialized war crimes units have convicted perpetrators of Syrian and other atrocities, though these are state prosecutions rather than victim-driven civil suits. Regional human rights courts in Europe, the Americas, and Africa hear cases against states and issue binding judgments. England’s courts have allowed negligence claims by foreign plaintiffs against British parent companies for overseas harms, achieving some corporate accountability through domestic tort law. The International Criminal Court prosecutes individuals for the gravest crimes, slowly and selectively. Nonjudicial tools include targeted sanctions such as the Global Magnitsky programs, truth commissions, and domestic prosecutions during political transitions. No single forum combines the ATS’s former reach of American discovery, jury trial, and punitive damages.