Alien Tort Statute

Summary: Supreme Court Rules in Jesner v. Arab Bank

By Chinmayi Sharma
Friday, April 27, 2018, 11:48 AM

This Tuesday, the Supreme Court held in Jesner et al. v. Arab Bank, PLC that the federal courts are not available to aliens in actions against foreign corporations.

In a 5-4 vote, with Justice Anthony Kennedy writing the majority opinion, the court affirmed the U.S. Court of Appeals for the Second Circuit’s dismissal of the case and held that aliens cannot bring suit under the Alien Tort Statute (ATS) against foreign corporations.

Procedural History

Petitioners originally brought five suits under the ATS in the Eastern District of New York between 2004 and 2010 on behalf of murdered and injured victims of terrorist acts that, plaintiffs contend, were knowingly funded by defendant, Arab Bank, PLC to the tune of several million dollars. A majority of the plaintiffs—about 6,000—are foreign nationals. They sought recovery for the deaths caused in part by the bank’s willingness to accept donations, maintain accounts, and transfer funds on behalf of and to known terrorist groups.

Arab Bank is a Jordanian financial institution with international branches, including one in New York. Petitioners originally brought action against the corporation to hold it accountable for the actions of its human agents, including the chairman, the board, and other high-ranking employees. The complaint alleged that knowing abetters of terrorist activities by Hamas and other groups violate human rights protections guaranteed by international law.

While many of the terrorist attacks in question occurred in the Middle East between 1995 and 2005, some of the funds used to finance the alleged terrorist accounts went through the New York office’s Clearing House Interbank Payment System (CHIPS), in transactions involving U.S. dollars as a primary or intermediary currency. Additionally, plaintiffs allege that the New York branch of Arab Bank was used to launder money for a Texas-based nonprofit suspected of supporting Hamas, the Holy Land Foundation for Relief and Development (HLF).

The district court dismissed the claims, holding as a matter of statutory interpretation and Second Circuit precedent that the ATS prohibited corporate liability. The Second Circuit affirmed this decision.

Petitioners filed for certiorari on the grounds that the district court decision relied on the Second Circuit’s ruling in Kiobel v. Royal Dutch Petroleum, which categorically precluded ATS suits against corporations but was affirmed by the Supreme Court on different grounds—leaving open the question of whether the ATS allows for corporate liability. The Supreme Court granted cert on April 3, 2017. By the time the court took up the case, the litigation had been ongoing for thirteen years.

The Supreme Court’s Opinion

The justices produced an array of fractured opinions. Writing for the majority, Kennedy, joined by Chief Justice John Roberts and Justices Clarence Thomas, Samuel Alito and Neil Gorsuch denied recognizing corporate liability on the basis of the judicial restraint called for by the language, purpose, and history of the ATS. Roberts and Thomas also joined Kennedy in a plurality opinion basing the decision on the application of the two-part test established in Sosa v. Alvarez-Machain (described below), which counsels against creating new corporate liability. Alito concurs in part, rather than applying the Sosa test, argues that separation of powers makes judicial creation of a new liability unconstitutional if it does not expressly serve the purpose of the ATS. Gorsuch also concurs in part and does not reach the Sosa test, writing that Article III diversity requirements and the original meaning of the ATS require the dismissal of the case at hand.

Majority Opinion: History of the ATS

Kennedy contextualizes the motivation for the ATS by describing the shortcoming of the U.S. Articles of Confederation in holding private parties accountable for harms to foreigners caused by treaty or international law violations. Accordingly, the founders afforded federal courts the power to adjudicate cases and controversies between Americans and foreigners in Article III, and implemented this power in the passage of the Judiciary Act of 1789.

The ATS was enacted as a part of the Judiciary Act to give district courts “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.” Kennedy cites the Supreme Court’s holding that the statute is “strictly jurisdictional” and does not give rise to any new causes of actions not already available under international law. It was originally enacted against the backdrop of an international common law understanding of obligations between individuals, as opposed to state obligations to other states and people, which at the time were limited to “violation of safe conducts, infringements of the rights of ambassadors, and piracy.”

It was against this background that the Supreme Court held in Sosa that Congress drafted the ATS “to furnish jurisdiction for a relatively modest set of actions alleging violations of the law of nations.” The goal was to avoid foreign entanglements by affording aliens a right to a remedy when the unavailability of one might provoke a foreign nation to hold the country accountable for the offending action instead. However, Kennedy goes on to explain that the statute was near-dormant for its first 190 years of its existence, until rising awareness of mass atrocities and events such as the Nuremberg Trials renewed its relevance.

Since the first case of this new era, Filartiga v. Pena-Irala, the Supreme Court has endeavored to identify boundaries to the ATS and balance the judicial role in this process against the political branches. In Filartiga, the family of a victim tortured and murdered by Paraguayan police officers filed suit against one of the officers residing in New York. Congress addressed doubts regarding the court’s jurisdiction over the case in the middle of the trial by enacting the Torture Victim Protection Act of 1991, which incorporated international law in creating a clear statutory cause of action for foreign plaintiffs in torture cases and extrajudicial killings. Kennedy states that the passage of the TVPA ushered in a period of more frequent and complex ATS suits involving “large groups of foreign plaintiffs suing foreign corporations in the United States for alleged human-rights violations in other nations.”

 

In light of this and the introduction of statutory causes of action, the Supreme Court further restrained its role in creating or extending common law causes of action. Fifteen years ago, in Sosa, the court reaffirmed its ability to recognize common law causes of actions based on modern day international law—but it reversed both lower courts’ interpretations that the ATS creates a new cause of action for violation of international law. The opinion especially called out the pressing separation-of-powers and foreign-relations concerns in these cases, counseling for “vigilant doorkeeping.”

In an effort to further avoid overstepping the judiciary’s role, the Supreme Court then applied the presumption against extraterritoriality to the ATS. Five years ago, in Kiobel v. Royal Dutch Petroleum, the court held that aliens cannot bring suit in a United States court for conduct that occurred entirely, or primarily, in another country, unless the domestic conduct is sufficient to displace the strong presumption against extraterritoriality. Although the case also involved a foreign corporation, the court did not reach that issue on the merits, dismissing the case on the grounds that its conduct was extraterritorial.

In Kiobel, Kennedy wrote a one-paragraph concurrence to Roberts’s majority opinion, suggesting that one day the extraterritoriality of the ATS may require “further elaboration and explanation.” In Jesner, Kennedy endeavors to provide just that.

Plurality Opinion

Corporate Liability

Kennedy explains the plurality’s decision to reach the issue of corporate liability under the ATS—that is, instead of simply dismissing the case on the grounds that Arab Bank did not meet the standard for minimum contacts with the U.S. for the court to have jurisdiction—as a matter of efficiency. “[I]f there is no liability for Arab Bank, the length and costly litigation concerning whether corporate contacts like those alleged here suffice to impose liability would be pointless.” Even here, Kennedy points to the ongoing friction this suit has caused between the United States and Jordan as a motivation for resolving the issue completely.

A plurality of the court would proceed to apply the test it established in Sosa: (1) Can the plaintiff demonstrate that the alleged violation is “of a norm that is specific, universal, and obligatory?” (2) Even assuming that, would allowing the case to proceed under the ATS be a proper exercise of judicial discretion, in light of concerns over foreign affairs and separation of powers? Kennedy acknowledges that the “two inquiries inform each other and are, to that extent, not altogether discrete.”

The first prong of the test is informed by footnote 20 from Sosa, which writes that a “related consideration is whether international law extends the scope of liability for a violation of a given norm to the perpetrator being sued, if the defendant is a private actor such as a corporation or individual.” The plurality reads this to mean that international law must inform not only the conduct prohibited but also which subjects can be held accountable.

In addressing the first prong of the Sosa test—whether the norm violated is “specific, universal, and obligatory”—the court’s opinion soundly rejects petitioners’ proposition that the shift in international law after World War II to holding individuals accountable for human rights violations, not just nation-states, extends to corporations or other artificial entities. As support, Kennedy cites the fact that international criminal tribunals “often exclude corporations from their jurisdictional reach,” from the Nuremberg Tribunal to more modern venues. He distinguishes the International Convention for the Suppression of the Financing of Terrorism, an example put forth by petitioners to argue that international law does in fact hold corporations accountable, as a treaty requiring nations to create laws to hold corporations accountable—not as a law holding corporations accountable themselves. In support, he argues the U.S. satisfied its obligations under this treaty by creating private rights of action under the Anti-Terrorism Act, the Bank Secrecy Act, and the Terrorism Sanctions Regulations, none of which extend causes of action to aliens. The convention does not require countries “to hold corporations liable in common-law tort actions raising claims under international law.”

Kennedy’s opinion acknowledges both that domestic law holds corporations accountable for the actions of their agents, and that “the enormity of the offenses that can be committed against persons in violation of international human-rights protections” can show that corporations should be held liable for the crimes of their agents. But he writes that “the international community has not yet taken that step, at least in the specific, universal, and obligatory manner required by Sosa.”

Petitioners also argued that the question of corporate liability is not one of substantive principle requiring the interpretation of law and precedent, but rather a case-by-case question of remedies available. The plurality finds this “far from obvious,” but does not reach a conclusion. By finding that there is sufficient doubt under the first prong of the test, the court moves to the second prong of the test to determine whether the judiciary should defer to Congress and allow it to “determine in the first instance whether that universal norm has been recognized and, if so, whether it is prudent and necessary to direct its enforcement in suits under the ATS.”

In engaging the second prong of the Sosa test, the plurality highlights its deferential precedence in issues involving foreign affairs, along with Congress’s superior competence in navigating policy implications of new private causes of actions.

First, it acknowledges that its recent precedents have “cast doubt on the authority of courts to extend or create private causes of action even in the realm of domestic law.” In Correctional Services Corp. v. Malesko, a case declining to extend a Bivens remedy, the Supreme Court refused to extend the availability of Bivens actions to hold corporations accountable. As with the case at hand, the court saw the issue raised by Malesko as a complicated issue best left for Congress.

The opinion continues to explain that “the separation-of-powers concerns that counsel against courts creating private rights of action apply with particular force in the context of the ATS,” because the statute necessarily implicates foreign policy. The court does not hold that the judiciary may never recognize a new cause of action under the ATS, but does conclude it is inappropriate to do so here.

In further support of this conclusion, the court points to the TVPA, the only cause of action under the ATS created by Congress, which only extends to natural persons. The court finds that, in so limiting the statute’s scope, Congress already made the policy decision ex ante about the value of holding corporations accountable under the ATS, which “[a]bsent a compelling justification, courts should not deviate from that model.”

Petitioners try to distinguish the TVPA from the case at hand by arguing that Congress limited the statute to individuals because it was concerned that, given that the legal definitions of torture and extrajudicial killings require state action, any corporation sued would necessarily qualify for sovereign immunity as an agent of a foreign state. The court counters this argument by citing to Mohamad v. Palestinian Authority, which held that the TVPA excludes all corporations, not just sovereign ones, and by pointing out that plaintiffs can use non-statutory causes of actions against corporations as surrogate defendants to challenge the conduct of foreign governments.

The court then rejects petitioners’ argument that the Anti-Terrorism Act’s ability to hold corporations liable is evidence that there should be a common law cause of action against corporations as well. Instead, the plurality holds that the ATA further supports the court’s conclusion that ATS does not extend to corporations: Congress limited the former statute to nationals, and acknowledging for a common law cause of action would allow aliens to circumvent Congress’s express intention in excluding them under the ATA by simply bringing a suit under the ATS.

Finally, the court concludes it is not necessary to hold corporations liable to carry out the goals of the ATS. First, it suggests plaintiffs can bring suit in other ways, such as seeking to hold the individual agents of the corporation accountable. In fact, the plurality implies this would be more desirable than plaintiffs ignoring “human perpetrators and concentrat[ing] instead on multinational corporate entities.” Second, the court raises the concern that holding foreign corporations liable under the ATS would open the door for reciprocal action, placing U.S. corporations at risk of foreign lawsuits. It fears setting a precedent that discourages foreign investment, especially in developing countries with potential histories of alleged human-rights violations. Third, it underlines the fact that foreign corporations generate complicated legal questions that hinge on questions of policy and foreign affairs, such as what qualifies as minimum contact with the United States under Kiobel. These unique challenges counsel judicial restraint. The court explicitly chooses not to decide “whether these allegations [against Arab Bank] are sufficient to ‘touch and concern’ the United States under Kiobel.”

Potential Political Responses

The plurality then engages in a discussion of potential paths the political branches can take to address corporate liability under the ATS, if they feel compelled to do so.

First, Congress could open up liability under the ATS to corporations by statute, if it finds that doing so would best serve the goals of holding parties accountable for human rights violations. Second, if Congress is concerned about reciprocal action from foreign nations, it can limit liability to corporations that meet certain preconditions. Finally, Congress can find that corporate liability is only appropriate when management has been “actively complicit in the crime.” The court uses these three possible paths, as well as potential others, to support the proposition that the appropriate solution requires a delicate balance of policy interests best made by the political branches.

The opinion explains that because Congress and the president comprise “the branches most immediately responsive to, and accountable to, the electorate” any action they take in holding corporations accountable would have “special power and force.”

Alito’s Concurrence

Justice Alito wrote separately, concurring with the majority in the decision that the ATS does not allow suits against foreign corporations, but arguing that this is required not only by “judicial caution,” which the majority focuses on, but also the separation of powers.

Alito explains that the ATS was enacted at a time when federal courts recognized “general common law,” which courts later struck down in Erie R. Co. v. Tompkins. Therefore, in Sosa, the court limited its own ability to create federal common law to cases in which Congress had conferred on it the authority to do so, through statutes such as the ATS.

Alito is “not certain that Sosa was correctly decided”—but assuming it is controlling, the second step of the Sosa test requires the court to decline creating a cause of action that would foil Congress’s purpose in enacting the ATS: avoiding diplomatic strife. In doing so, the court would be “unconstitutionally usurping” the legislative branch’s authority. The opinion proceeds to point out various examples of ATS suits against foreign corporations provoking international strife, and notes that “when states do not object to this sort of corporate liability as a legal matter, they may be concerned about ATS suits against their corporations for political reasons.”

Alito rejects any judicial creation of liability unless there is affirmative evidence that the creation of that liability would “actively decrease diplomatic disputes.” In the absence of this evidence, such action would be unauthorized and therefore unconstitutional. The possibility that liability might theoretically ameliorate diplomatic tensions, as proposed by the dissent, does not meet this standard.

Gorsuch’s Concurrence

Gorsuch writes that the lawsuit should be dismissed for two fundamental reasons not touched on by the majority. He “would end ATS exceptionalism” and urges future courts to consider his justifications for judicial restraint before “taking up cases like this one.”

First, Gorsuch casts doubt on the holding in Sosa that allows judges to recognize common law causes of actions beyond the original three recognized under English common law as adopted by the American courts at the time the ATS was passed. He sees such undertakings by courts as going beyond judicial duty into policy-making, which is best and most appropriately left in the province of the political branches. As stated in Sosa, the ATS is no exception because it was meant to serve strictly jurisdictional purposes; in his view, “[a] statute that creates no new causes of action ... creates no new causes of action.”

But even if the court continues to uphold the Sosa framework for judicially created causes of action, it must still refuse to recognize a new cause of action in this case. Gorsuch agrees with the majority that if a cause of action has not been warranted in the hundreds of years since the statute’s passage, it is not warranted now. He calls for judges to decline such activism when doing so necessarily requires making judgments about “delicate questions of national affairs.”

Finally, he claims that Article III of the Constitution prevents the court from hearing this suit, because as a matter of interpretation, it requires there to be at least one domestic defendant for courts to hear a case. The diversity-of-citizenship requirement for federal cases cannot typically be satisfied by two sides consisting entirely of aliens, and Gorsuch argues the same standard is applicable for the ATS. As support, Gorsuch writes that other statutes enacted under the Judiciary Act alongside the ATS, which also involve civil cases “where an alien is a party,” have been interpreted to require diversity conditions to be met. He would limit the ATS to cases in which a foreign plaintiff holds a domestic defendant accountable under international law.

Thomas’s Concurrence

Justice Thomas joins the majority and plurality opinions in full, for “correctly applying our precedents” while also agreeing with Alito and Gorsuch’s concurrences. He supports Gorsuch’s call for future judicial restraint from creating new causes of action under the ATS and his conclusion that the ATS does not apply to suits between foreign plaintiffs and foreign defendants. He also echoes Alito’s concern that the court must exercise extra caution when creating a new cause of action would risk international strife.

Dissent

Justice Sonia Sotomayor, writing the dissent and joined by Justices Ruth Bader Ginsburg, Stephen Breyer and Elena Kagan, criticizes the court for “absolv[ing] corporations from responsibility under the ATS for conscience-shocking behavior.” She argues that nothing in the “corporate form in itself” raises foreign policy concerns distinct from foreign individuals in a manner that would require the court to immunize corporations.

 

First Prong of the Sosa Test

First, Sotomayor writes that the plurality misunderstands the nature of international law by asking whether there is “a specific, universal, and obligatory norm of liability for corporations,” and thereby misapplies the first step of Sosa. That first step focuses on the substantive norms of international law that prohibit certain conduct, such as slavery, genocide, and extrajudicial killing. Sotomayor argues that while Sosa requires there to be international consensus about the norm to be violated, there need not be such agreement to decide that corporations should be held liable.

Sotomayor points to the text of the statute to support her argument that there is a “distinction between prohibiting conduct and determining enforcement.” The language of the statute reads: “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.”

Sotomayor interprets “of the law of nations” to modify “violation,” not “civil action.” This leaves the door open for countries to establish appropriate means of enforcement as they see fit, rather than looking to international law.

Sotomayor rejects the court’s reading of footnote 20 in Sosa, interpreting it not as a charge to courts to look to international law in limiting liability against subjects, but as an acknowledgement that the violative nature of conduct can turn on the nature of the perpetrating party as either an independent actor or an actor on behalf of the state. Therefore, she believes the court must ask under the first prong of Sosa whether the norm of financing terrorist conduct is ubiquitously prohibited. Then it must ask whether there is “sufficient consensus” that the defendant being sued can be alleged with violating that norm. Here, she characterizes the Arab Bank as independent of the state at issue—so the bank qualifies under international legal standards as a party prohibited from financing terrorism.

Having established that the norm underlying the ATS claim is supported by international law, Sotomayor writes that it is a matter left to states whether the norm should be enforced by holding accountable the responsible individual or the corporate entity through which the individual worked. She criticizes the plurality’s emphasis on the absence of criminal tribunals holding corporations accountable,  as there has been no such corollary tribunal to hold natural persons civilly liable. Instead, looking to states’ collectively and individual enforcement actions “makes clear that corporations are subject to certain obligations under international law.” Accordingly, Sotomayor lists several examples including the United States Military Tribunal’s prosecution of corporate executives and the International Criminal Tribunal for Rwanda’s finding three that non-natural entities were responsible for genocide.

The dissent also calls attention to International Convention for the Suppression of the Financing of Terrorism, which requires nations to hold individuals and corporations accountable for funding terrorist activities under domestic law. By prohibiting this conduct by corporations and allowing states to design their own enforcement schemes, the convention maintains the possibility for countries to allow foreign individuals to sue foreign corporations, at the discretion of the state.

Having rejected the plurality’s theory that the court must look to international law to determine whether corporations can be held liable, Sotomayor proceeds to interpret the statute itself to assess whether the ATS can, under a plain language reading, hold corporations liable. She concludes that it can: The statute’s language confers jurisdiction for civil tort action, under which corporations have been historically held accountable. What’s more, the statute limits the class of potential plaintiffs to aliens, but includes no such limiting language about classes of defendants.

In confronting Gorsuch’s concurrence, Sotomayor first rebukes the claim that Sosa overstepped in holding that courts were authorized to recognize private causes of action based on the modern law of nations. She argues that if Congress sought to limit the available torts under the ATS to “violations of safe conduct, assaults against ambassadors, [and] piracy,” then lawmakers could have done so by adding that limit in plain language to the statute. Instead, Congress created causes of action arising under the “law of nations” and “treaties of United States”—which suggests torts as they develop over time, not limited to the three causes of actions described by Gorsuch as available at the time. Therefore, she urges, we must ask: “[W]ho are today’s pirates?”

Second, Sotomayor rejects Gorsuch’s theory that Article III requires a domestic defendant in an ATS suit. As a legal matter, she argues, Sosa foreclosed this argument by allowing claims under “federal common law,” which is not similarly limited by Article III’s diversity requirements as a matter of pragmatism. She references the logic in Saeri v. Rio Tinto, a case in which the Supreme Court concluded that federal courts have Article III jurisdiction to hear ATS cases between aliens and that the first Congress’s intentions could not be interpreted to remove federal courts’ capacity to enforce international norms because of formalistic limits. Therefore, the Saeri court held there are “limited enclaves in which federal courts may derive from substantive law in a common law way.” Further, Sotomayor contends, “our Nation has an interest not only in providing a remedy when our own citizens commit law of nations violations, but also in preventing our Nation from serving as a safe harbor for today’s pirates.”

Second Prong of the Sosa Test

In assessing the plurality’s analysis under the second prong of the Sosa test, Sotomayor argues that there is nothing inherent about corporations that justifies precluding all future ATS suits against them.

She believes the court mischaracterizes “the relatively minor connection between the terrorist attacks at issue in this case and the alleged conduct in the United States” as evidence of the perils of extending liability to foreign corporations, when in fact that connection merely goes to an assessment of whether the claims “touch and concern the territory of the United States ... with sufficient force” to overcome the presumption against extraterritoriality. Arab Bank’s complaint that it was being sued as a surrogate for the acts of foreign governments is not an argument against corporate liability, but simply an expression of frustration over its liability for aiding and abetting terrorism. Sotomayor further states that even Jordan, with whom the court argues the United States risked souring relations, was not worried about corporate liability. Rather, Jordan felt the suit was an affront to sovereignty because of its extraterritorial nature, not because the defendant was a corporation.

In this way, Sotomayor brushes aside the court’s conclusion under the second prong of the Sosa test as a misapplied concern about extraterritoriality, not corporate liability. Sotomayor would have remanded the suit to the Second Circuit to address whether any existing doctrine can serve to hold Arab Bank accountable. The Second Circuit could have concluded that the defendant’s minimum contacts with the United States were insufficient and thereby dismissed the case for lack of personal jurisdiction. Or, the Second Circuit could have dismissed the ATS suit if the plaintiff had not exhausted the remedies available in her domestic forum, on forum non conveniens grounds, for reasons of international comity. These options would allow the judicial branch to address concerns about extraterritoriality or inadvertent incitement of geopolitical strife without foreclosing all ATS actions against corporations. The majority, she argues, “prefers to use a sledgehammer to crack a nut.”

Addressing the court’s concerns regarding deference to the political branches, Sotomayor suggests that both the executive and legislative branches have expressed support for holding corporations liable in various reports and briefs weighing in on related matters.

She then argues that the court over-relies on the TVPA to conclude that by limiting liability in cases of torture and extrajudicial killing to natural persons, Congress meant to foreclose all international law ATS liability for corporations. She argues that the TVPA first makes legal action available to all persons—not just all aliens, like the ATS—and it uses the word “individuals” in limiting defendants, which the ATS does not. Accordingly, she points to precedents in Mohamed and congressional reports that state the TVPA was meant to augment, not curtail, the ATS.

Next, Sotomayor responds to the court’s conclusion that the availability of corporate liability for terrorism suits brought by U.S. nationals under the ATA suggests Congress did not intend for corporate liability under the ATS. First, she argues that the ATA confirms that “Congress exercises its judgment as to the appropriateness of corporate liability on a norm-by-norm basis, and that courts should do the same when considering whether to permit causes of action corporations under the ATS.” Second, she rejects the criticism that allowing corporate liability would allow aliens to bypass the limitations of the ATA by bringing suit under the ATS: litigants have the ability to bring both suits simultaneously, and have even done so in the case before the court. Instead, the ATA should be read as extending to U.S. nationals the legal avenues already available to aliens under the ATS. The statutes are not incompatible but complementary.

Although the court asserts that plaintiffs were unable to show that corporate liability is essential to the ATS, Sotomayor responds that the court has never previously required a plaintiff to show the ATS is the exclusive means of recovery to succeed. But even so, she believes there is a strong argument that leaving open the possibility of corporate liability in this or a future case may be the best way to serve Congress’s intent in enacting the ATS. She describes the uniquely powerful and problematic role corporations play in terrorism today—especially as financiers of crimes in exchange for potential profit—portending that an inability to hold corporations accountable for international violations would make impossible the statute’s goal of avoiding international conflict by providing foreign nationals with due remedies.

Sotomayor concludes by calling attention to the stark contrast of a corporation’s potential domestic liability for wrongdoing but its complete immunity from “human rights abuses [abroad], however egregious they may be.”