concurring.
I agree with the majority opinion, except for Part 11(A), in which the majority discusses the Alien Tort Claims Act. As to that Act, I agree with the majority that material factual disputes exist regarding plaintiffs’ claims for forced labor used in connection with the Yadana Pipeline Project. I also agree with the majority that if plaintiffs prove their allegations, Unocal may be held liable under the Act for the use of forced labor as a part of the project. Where I differ from my colleagues is principally with respect to the standard of third-party liability under which Unocal may be held legally responsible for the human rights violations alleged. I do not agree that the question whether Unocal may be held liable in tort for the Myanmar military’s alleged human rights violations should be resolved, as the majority holds, by applying a recently-promulgated international criminal law aiding-and-abetting standard that permits imposition of liability for the lending of moral support. In fact, I do not agree that the question of Unocal’s tort liability should be decided by applying any international law test at all. Rather, in my view, the ancillary legal question of Unocal’s third-party tort liability should be resolved by applying general federal common law tort principles, such as agency, joint venture, or reckless disregard. I also believe that there is no reason to discuss the doctrine of jus cogens in this case. Because the underlying conduct alleged constitutes a violation of customary international law, the violation was allegedly committed by a governmental entity, and Unocal’s liability, if any, is derivative of that government entity’s, jus cogens is irrelevant to any issue before us. Assuming the allegations to be true, the fact that the underlying conduct violated customary international law is sufficient to support liability not only on the part of the governmental actor, but also on the part of a third party whose liability is derivative thereof.
1. Forced Labor As A Violation of the Law of Nations
In order to bring an action under the Alien Tort Claims Act, an alien plaintiff must allege a tort committed in violation of the law of nations. Hilao v. Estate of Marcos, 25 F.3d 1467, 1475 (9th Cir.1994). I agree with the majority that the plaintiffs have alleged the requisite international law violation, and that a genuine issue of material fact exists regarding whether forced labor was used by the Myanmar government in connection with the Yadana Project. Because the majority opinion thoroughly sets forth the plaintiffs’ serious allegations, and the evidence supporting those allegations, I do not repeat them here.
There can be little doubt that the use of forced labor violates widely-held international legal norms. Forced labor is banned by the Universal Declaration of Human Rights, the International Covenant of Civil and Political Rights, and the International Covenant of Economic, Social and Political Rights. See Universal Declaration of Human Rights, G.A. Res. 217(A)III (1948); International Covenant on Civil and Political Rights, art. 22, 999 U.N.T.S. 171, 173-74; 6 I.L.M. 368; International Covenant of Economic, Social and Political *964Rights, art. 8, 993 U.N.T.S. 3, 4; 6 I.L.M. 360. Forced labor was listed as a war crime in the charter of the Nuremberg Tribunal. See Agreement for the Prosecution and Punishment of the Major War Criminals of the European Axis, and Charter of the International Military Tribunal, Aug. 8, 1945, art. 6, 82 U.N.T.S. 280. Moreover, it is banned by two of the most widely-adopted international labor conventions. Convention Concerning the Abolition of Forced Labour (No. 105), June 25, 1957, 320 U.N.T.S. 291; Convention Concerning Forced or Compulsory Labour (No. 29), June 28, 1930, 39 U.N.T.S. 55. In light of these legal authorities, the allegations of forced labor practices, if true, constitute a violation of customary international law and, in any event, are sufficient to confer jurisdiction under the ATCA.
In contrast, the majority states that plaintiffs have alleged the necessary international law violation because forced labor is a modern variant of slavery, which is a jus cogens or “peremptory norm” of international law.1 In fact, whether or not forced labor is a modern variant of slavery is of no legal consequence in this case, because there is no requirement that plaintiffs state a jus cogens violation in order to obtain jurisdiction under the ATCA. It is true that a cause of action against non-state actors for conduct in which they engage directly exists only for acts that constitute jits cogens violations and that other conduct of private parties that would violate international law if engaged in by a governmental entity is not actionable under the ATCA. See Kadic v. Karadzic, 70 F.3d 232, 240 (2d Cir.1995). Here, however, if Unocal is held liable, it will be because the Myanmar military committed the illegal acts and Unocal is determined to be legally responsible for that governmental conduct under a theory of third-party liability — not because Unocal itself engaged in acts transgressing international law. Because the violations of customary international law, if they occurred, were committed by a governmental agency, third-party liability may attach regardless of whether the international law violated is jus cogens.2 Thus, I see no need to discuss whether forced labor is a modern variant of slavery, which would render it a jus cogens norm, or even whether the prohibition on forced labor is itself a jus cogens norm, which it may well be. See Princz v. Fed. Republic of Germany, 26 F.3d 1166, 1179-81 (D.C.Cir.1994) (Wald, J., dissenting on other grounds). The well-established principle that forced labor practices violate customary international law is sufficient in itself to confer jurisdiction in this ease with respect to all parties, jus cogens or not.
2. The Appropriate Source of Law for Determining Third-Party Liability
If the plaintiffs can prove their allegations that the Myanmar military instituted a policy of forced labor, they would satisfy the Alien Tort Claims Act requirement of *965a violation of the law of nations. Then, in order to prevail on its claims against Unocal, plaintiffs would have to prove that the private entity may be held legally responsible for the Myanmar military’s human rights violations. The latter requirement raises important questions of first impression: Under what circumstances may a private entity doing business abroad be held accountable in federal court for international law violations committed by the host government in connection with the business activities of the private entity; and to what body of law do we look in order to determine the answer? Logically, it is necessary to consider the second question first. In my view, the answer is that we look to traditional civil tort principles embodied in federal common law, rather than to evolving standards of international law, such as a nascent criminal law doctrine recently adopted by an ad hoc international criminal tribunal.
Plaintiffs allege that Unocal should be held hable for Myanmar’s forced labor actions with respect to the pipeline under a number of international law theories, as well as under several theories based on federal common law principles. The text of the Alien Tort Claims Act states only that federal courts have jurisdiction over torts constituting a violation of “the law of nations.” It is thus clear from the face of the statute that international law applies to determine whether a violation has occurred. The statute is silent, however, as to what body of law applies to ancillary issues that may arise, such as whether a third party may be held liable in tort for a governmental entity’s violation of the law of nations. The majority elects to apply international law principles to resolve such issues. I strongly disagree. I believe that we are required to look to federal common law to resolve ancillary, legal issues that arise in ATCA cases.
Following Erie Railroad Co. v. Tompkins, 304 U.S. 64, 58 S.Ct. 817, 82 L.Ed. 1188 (1938), federal courts ordinarily apply federal common law in limited circumstances, usually when authorized to do so by Congress. However, actions involving international relations constitute one category of cases in which federal common law is frequently applied. The Supreme Court has stated that even without congressional authority to develop federal common law, the federal courts should apply such law “in such narrow areas as those concerned with the rights and obligations of the United States, interstate and international disputes implicating the conflicting rights of States or our relations with foreign nations, and admiralty cases.” Texas Industries, Inc. v. Radcliff Materials, Inc., 451 U.S. 630, 641, 101 S.Ct. 2061, 68 L.Ed.2d 500 (1981); see also Banco Nacional de Cuba v. Sabbatino, 376 U.S. 398, 425, 84 S.Ct. 923, 11 L.Ed.2d 804 (1964) (“It seems fair to assume that the Court did not have rules like the act of state doctrine in mind when it decided Erie R. Co. v. Tompkins.”). Because Alien Tort Claims Act cases involve the violation of international law, they almost always “implicate] ... our relations with foreign nations.” Texas Industries, 451 U.S. at 641, 101 S.Ct. 2061.3 There are thus unique federal interests involved in Alien Tort Claims Act cases that support the creation of a uniform body of federal common law to facilitate the implementation of such claims.
There is another reason why the application of federal common law is appropriate here: we are required to resolve issues *966ancillary to a cause of action created by Congress. The Supreme Court has stated that in such cases, courts should apply federal common law “to fill the interstices of federal legislation.” United States v. Kimbell Foods, Inc., 440 U.S. 715, 727, 99 S.Ct. 1448, 59 L.Ed.2d 711 (1979). In this regard, federal common law is applicable where courts are required to implement the policies underlying a federal statute by fashioning appropriate remedies. Illinois v. City of Milwaukee, Wise., 406 U.S. 91, 100-04, 92 S.Ct. 1385, 31 L.Ed.2d 712 (1972) (holding that federal courts may fashion federal common law remedies to implement the policies of federal water pollution statutes, because interstate navigable waters are inherently a matter of federal concern, and the federal legislation did not address the specific legal issue presented.); see County of Oneida, N.Y. v. Oneida Indian Nation of New York State, 470 U.S. 226, 237, 105 S.Ct. 1245, 84 L.Ed.2d 169 (1985) (applying federal common law to a remedial question where the federal Non-Intercourse Act failed to address what legal standard to apply). It is precisely in order to implement the policies underlying Congress’s decision to make the violation of international law a federal tort, that it is necessary to flesh out the statute and apply federal common law; here, we must do so in order to fashion a remedy with respect to the direct or indirect involvement of third parties in the commission of the underlying tort.4
Next, the question of when third-party liability arises is a straightforward legal matter that federal courts routinely resolve using common law principles. See cases cited Part 3, infra. It is not an issue of such rarity, so seldom broached and so puzzling that our domestic law offers inadequate guidance and we are compelled to look elsewhere. The fact that some of the acts at issue here may have taken place abroad does not militate in favor of applying international law; transnational matters are litigated in federal court, using federal legal standards, more and more frequently as the pace of globalization grows ever more rapid. Nor is there any reason to apply international law to the question of third-party liability simply because international law applies to the substantive violation; as discussed above, federal common law is properly invoked when the statute at issue leaves an ancillary question unanswered, regardless of the nature of the statute. In short, federal common law principles provide the traditional and time-tested method of filling in the interstices and resolving the type of ancillary legal questions presented by this case.
In my view, courts should not substitute international law principles for established federal common law or other domestic law principles, as the majority does here, unless a statute mandates that substitution, or other exceptional circumstances exist. Examples of when the substitution of international law is appropriate include interpreting the substantive provisions of the Torture Victims Protection Act, Pub.L. No. 102-256, 106 Stat. 73 (1992) (codified at 28 U.S.C. § 1350), certain provisions of the Foreign Sovereign Immunities Act, 28 U.S.C. § 1330 et seq., and the substantive component of the ATCA. In those cases, the statutes make it plain that certain provisions require the application of international law. In other instances, I believe it prudent to follow the general rules es*967tablished by the Supreme Court regarding the use. of federal common law. It is important to recognize that there is a distinction between substituting international law for federal common law and making proper use of international law as part of federal common law. Employing federal common law does not force courts to ignore a constructive or helpful rule adopted under international law, because in appropriate circumstances federal common law incorporates relevant principles of international law. The Paquete Habana, 175 U.S. 677, 700, 20 S.Ct. 290, 44 L.Ed. 320 (1900); see also n. 7, infra. Thus, the benefits of the vast experience embodied in federal common law as well as any useful international law principles are obtained when we employ the traditional common law approach ordinarily followed by federal courts. Those benefits are lost, however, when we substitute for the wide body of federal authority and reasoning, as the majority does here, an undeveloped principle of international law promulgated by a recently-constituted ad hoc international tribunal.5
Almost all of the factors that we are required to consider as part of a choice-of-law inquiry militate in favor of determining that the proper law to apply here is the federal common law.6 First, “ease in the determination and application of the law to be applied” is furthered by applying a well-developed body of law, as opposed to a standard announced in a criminal case only recently decided by an ad hoc international tribunal. Similarly, “certainty, predictability and uniformity of result” are more likely to be achieved when there exists extensive precedent upon which to draw, and the state of the law does not depend on the future decisions of some as-yet unformed international tribunal established to deal with other unique regional conflicts. Additionally, although the “justified expectations” of potential parties may be limited in the sense that no direct precedent exists on the question of third-party liability in ATCA cases, the federal common law principles of joint liability, agency, and reckless disregard that we regularly apply in other contexts are generally well-known. In contrast, the international law regarding third party “moral support” is of very recent origin, and our selection of that law would not lead to settled expectations in future cases; for, the standard may well change dramatically if and when it is applied by a different ad hoc tribunal appointed by future representatives of the nations that compose the General Assembly of the United Nations. Next, as noted earlier, the policy of the Alien Tort Claims Act is “to establish[ ] á *968federal forum where courts may fashion domestic common law remedies to give effect to violations of customary international law.” Abebe-Jira v. Negewo, 72 F.3d at 848. Thus, the “relevant polic[y] of the forum” is to apply federal common law remedies such as the imposition of third-party liability in the case of violations of customary international law. Finally, “the basic polic[y] underlying the particular field of law” is to provide an appropriate tort remedy for certain international law violations. The application of third-party liability standards generally applicable to tort cases directly furthers the basic policy of using tort law to redress international wrongs, whereas the application of international criminal law doctrines does not advance that objective. The two remaining choice-of-law factors are neutral, at the least, and certainly do not support rejecting the use of federal common law.7
Moreover, although much of international human rights law has developed in the criminal context, as the majority notes, the question of how to establish third-party liability is not in any way unique to human rights cases. The fact that the substantive violation involves international prohibitions on forced labor rather than a more traditional tort does not present any different concerns with respect to the determination of third-party liability. I thus see no reason to look to international criminal law doctrines for a civil liability standard when a substantial body of federal common law already exists regarding third-party liability generally. In sum, because Supreme Court precedent concerning the application of federal common law dictates its application here, and because the accepted choice of law factors overwhelmingly militate in favor of applying federal common law, I would derive a third-party liability standard for ATCA cases from that body of law.
Finally, I would note that the majority’s disclaimer in its opinion that its legal conclusion regarding the applicability of international law rather than federal common law “is based on the record before it”, and that in cases with other facts federal common law may apply, see maj. op., n. 25, serves no apparent purpose other than to attempt to distance the majority from its choice of international law. Indeed the footnote undermines the opinion’s fundamental holding. All appellate decisions are based on the record before the court. More important, in all cases in which a third party is alleged to be legally responsible for the acts of a host government, third part liability determinations must be based on a principled choice of law. We must decide whether either international law or federal common law is applicable to the category of cases at issue here. The choice of law in such cases does not depend on the facts of the particular case, nor does it vary with the particular circumstances of the case. A binding legal rule must apply, or to put it differently, a controlling legal principle must govern, the legal question involved, regardless of the particular facts of a case. Either international law applies to third-party liability issues in ACTA cases or it doesn’t. Either the ancillary question of whether an American corporate entity may be held liable under ACTA for the conduct of a host government is governed by federal common law or it isn’t. What varies from case to case is not the question of the governing law, but whether liability attaches in the particular instance. That determination is based on the facts and circumstances of *969the case and that determination must be made regardless of whether international law or federal common law has been held to be applicable when resolving the issue of third-party liability. Thus, I reiterate that, unlike the majority, I would hold that the ancillary issue of third-party liability in ACTA cases must be decided as a matter of federal.common law. I would not have the choice of law depend on the facts of the particular case.
S. Application of Federal Common Law
Having determined that ancillary legal issues in ATCA cases are to be resolved in accordance with federal common law, the question remains, for me, as to the proper federal common law rule for. third-party liability in this case. Federal common law has developed over time a number of principles under which courts determine whether third parties may be held liable for the wrongful acts of others. Specifically, as will be discussed infra, the principles of joint venture, agency, and reckless disregard have all been applied across a wide range of torts and other legal wrongs, and the overwhelming weight of federal authority supports their application here. .
Before turning to the application of the three federal common law theories to the instant case, it is necessary first to consider briefly whether the international law principle adopted by the majority may be applicable as part of the federal common law. It plainly is not. As noted earlier, international law principles may, under appropriate circumstances, become a part of the federal common law. Specifically, when an international legal principle achieves sufficient international acceptance that it constitutes customary international law, it also becomes part of the federal common law. - Filartiga v. Pena-Irala, 630 F.2d 876, 881 (2d Cir.1980).8 However, the Yugoslav Tribunal’s “moral support” standard is far from such a settled rule. As I observed earlier, it is a novel standard that has been applied by just two ad hoc international tribunals. It does not constitute customary international law, and thus we are not free to apply it as part of federal common law.
Moreover, even if it were possible for this court to determine that the Yugoslav Tribunal’s novel criminal standard constitutes a part of the federal common law, I would strongly doubt the wisdom of using that rule to override the well-established federal common law tort principles that would otherwise be applicable to resolving third-party tort liability questions. The Yugoslav Tribunal’s standard provides that an individual may be liable for “practical assistance, encouragement, or moral support which has a substantial effect on the perpetration of the crime.” Prosecutor v. Furundzija, IT-95-17/1-T (Dec. 10, 1998), reprinted in 38 I.L.M. 317, ¶ 235 (1999). It is, in my view, far too uncertain and inchoate a rule for us to adopt without *970further elaboration as to its scope by international jurists. Although it is of some comfort that the majority considers “moral support” to be equivalent to “encouragement” in domestic tort law, it is nevertheless far from clear what the practical implications would be of adopting the standard recently announced by the ad hoc tribunal on war crimes in the former Yugoslavia.9 Members of a future ad hoc tribunal elected by representatives of all of the nations that may then belong to the United Nations General Assembly might well define the term quite differently than does the majority here. Thus, the unintended consequences of adopting the ad hoc tribunal’s “moral support” standard may be significant.10
Returning to the three federal common law theories of third-party liability, plaintiffs alleged all three in their complaint, all of the theories are well-established in the federal common law, and disputed questions of fact exist with respect to each. Thus, in my view, plaintiffs are entitled to proceed to trial on the basis of each of the three theories. I address the basis for each theory in federal common law, as well as the evidence in the record supporting each:
a. Joint Venture Liability
It is well-established as a federal common law principle that a member of a joint venture is liable for the acts of its co-venturers. Federal courts freely invoke this principle of liability when called upon to apply federal common law in a variety of contexts. See, e.g., Davidson v. Enstar Corp., 848 F.2d 574, 577-78 (5th Cir.1988) (applying federal common law of joint liability rather than the idiosyncratic Louisiana law of joint liability in determining whether a relationship constituted a joint venture for purposes of the Longshore and Harbor Workers’ Compensation Act); United States v. United Pacific Insurance Co., 545 F.2d 1381, 1382-83 (4th Cir.1976) (applying the federal common law of joint venturer liability in interpreting the Miller Act, 40 U.S.C. § 270a-d). Moreover, “different jurisdictions generally adopt the same criterion for the establishment of a joint venture.” United States v. USX Corp., 68 F.3d 811, 826 n. 30 (3d Cir.1995).11
*971The principle that a member of a joint venture is liable for the torts of its co-venturer is well-established in international law and in other national legal systems. International legal materials frequently refer to the principle of joint liability for co-venturers. See, e.g., United Nations Convention On the Law of the Sea, Art. 139, Oct. 21, 1982, 21 I.L.M. 1245, 1293 (establishing principle of joint liability in international maritime law for parties acting jointly in maritime ventures); Convention on International Liability for Damage Caused by Space Objects, Mar. 29, 1972, 961 U.N.T.S. 187 (establishing joint liability principles to harms caused by parties launching objects into space); see also John E. Noyes & Brian D. Smith, State Responsibility and the Principle of Joint and Several Liability, 13 Yale J. Int’l. Law 225, 249 (1988) (describing joint and several liability for co-venturers’ actions as a general principle of international law). The status of joint liability as a general principle of law is supported not only by international law sources but also by the fact that it is fundamental to “major legal systems.” See, e.g., N.Y. PARTNERSHIP Law § 24 (McKinney 2002); Buckley v. Chadwick, 45 Cal.2d 183, 190, 288 P.2d 12 (1955); Caron v. Lynn Sand & Stone Co., 270 Mass. 340, 346, 170 N.E. 77 (1930); 2 Laws and Regulations of the People’s Republio of China 71 (1984) (Chinese joint venture statute); AIB Group (UK) Pic. v. Martin, 2001 U.K.H.L. 63 (United Kingdom joint venture law).
The body of international law described above serves to confirm my view that federal common law regarding the liability of joint venturers applies in the Alien Tort Claims Act context in the same manner and to the same extent as it does in construing other federal statutes. I would therefore. hold that plaintiffs may recover on a federal common law theory of joint liability if they can prove both that the forced labor violations occurred and that Unocal was a co-venturer with the Myanmar military, which perpetrated the violations.
As discussed above, there exists a question of fact requiring trial regarding the occurrence of the forced labor violations. There also exists a question of fact regarding whether Unocal and the Myanmar military were co-venturers. The corporate entity that oversaw the gas exploration project consisted of four partners: Unocal, Total, the Myanmar government (which is a military regime and thus indistinguishable from the military), and a Thai corporation. Thus, contrary to Unocal’s contentions, the evidence supports more than the conclusion that Unocal simply chose to invest in a project that happened to take place in a nation in which human rights abuses were widespread. Rather, a reasonable jury could conclude that Unocal freely elected to participate in a profit-making venture in conjunction with an oppressive military regime — a regime that had a lengthy record of instituting forced labor, including forced child labor.
Unocal contends that the Myanmar regime was a partner only in the offshore drilling portion of the Yadana project and not in the pipeline Construction portion of the project. The company argues that a Total affiliate and a Unocal subsidiary created a pipeline construction corporation (called the Moattama Gas Transportation Company, or “MGTC”) which was independent of both the joint venture and the military. A factual dispute exists with respect to this contention. Significantly, however, one of Unocal’s business manag*972ers stated that “the [Yadana] project is an entirety ... although there may appear to be two different businesses ... this is an illusion.” There is substantial evidence in the record that MGTC was the alter ego of the joint venture, in which case Unocal would be responsible for torts committed by its co-venturer, the Myanmar military, in the course of the pipeline construction company’s activities. Plaintiffs contend that despite the existence of MGTC, the only reasonable reading of the contract forming the joint venture is that the joint venture is also responsible for the pipeline construction. Plaintiffs also argue that MGTC is a shell corporation because it maintained no independent offices, was under-capitalized, and relied only on the employees of the joint venture. Finally, evidence in the record states that Unocal would share revenues and costs of both the drilling and transportation components of the Yadana project. In view of the above, I believe that plaintiffs ought to proceed to trial on their claim of joint venture liability.
b. Agency Liability
Plaintiffs contend that Unocal may also be held liable for the acts of the Myanmar military because the military acted as the company’s agent. The theory of agency liability is also well-supported in the federal common law. The Supreme Court has observed in the context of the Copyright Act that “when we have concluded that Congress intended terms such as ‘employee,’ ‘employer,’ and ‘scope of employment’ to be understood in light of agency law, we have relied on the general common law of agency, rather than on the law of any particular State, to give meaning to these terms.” Community for Creative Nonviolence v. Reid, 490 U.S. 730, 740, 109 S.Ct. 2166, 104 L.Ed.2d 811 (1989); see also Gleason v. Seaboard Air Line Ry. Co., 278 U.S. 349, 356, 49 S.Ct. 161, 73 L.Ed. 415 (1929) (“[F]ew doctrines of the law are more firmly established or more in harmony with accepted notions of social policy than that of the liability of the principal without fault of his own.”)
Plaintiffs’ theory of agency liability is consistent with the substantial federal common law of agency developed in the context of the Labor-Management Relations Act and ERISA. See, e.g., Anderson v. International Union, United Plant Guard Workers of America, 150 F.3d 590, 592-93 (6th Cir.1998) (addressing an agency issue under ERISA and holding that “we are guided by the law of agency as developed and interpreted as a matter of federal common law.”); National Football Scouting, Inc. v. Continental Assurance Co., 931 F.2d 646, 648 (10th Cir.1991) (examining whether “under the federal common law of agency” an agent of a plan fiduciary was acting within his actual or apparent authority). That federal common law should govern plaintiffs’ claim that the Myanmar military acted as Unocal’s agent.
The general principles of the federal common law of agency have been formulated largely based on the Restatement of Agency. Moriarty v. Glueckert Funeral Home, Ltd., 155 F.3d 859, 866 n. 15 (7th Cir.1998); Cilecek v. Inova Health Sys. Servs., 115 F.3d 256, 259-60 (4th Cir.1997). Under those general principles, an agency relationship may be express or implied; in addition, a jury may infer from the factual circumstances that apparent agency authority exists. See Hasbrouck v. Sheet Metal Workers Local 232, 586 F.2d 691, 693 (9th Cir.1978); see also 3 AM. Jur. 2d Agenoy § 21 (1986) (“The manner in which the parties designate the relationship is not controlling, and if an act done by one person in behalf of another is in its essential nature one of agency, the one is the agent of such other notwithstanding he is not so called.” (footnote omitted)).
*973As is true of joint liability principles, agency liability principles are well-established in international law. They are frequently discussed in international legal materials. See, e.g., Case Concerning the Barcelona Traction Power & Light Co. (Bel. v. Spain), 1970 I.C.J. 3, 215 (discussing principal-agent liability as applicable to commercial relationships generally); Application for Review of Judgment No. 333 of the U.N. Administrative Tribunal (Yakimetz Case), May 27, 1987 (applying principal-agent liability to acts of United Nations employees).12 Principal-agent liability is also widely adopted by civil law and other common law systems. See, e.g., Bazley v. Curry, 2 S.C.R. LEXIS 134 (1999) (Canadian Supreme Court statement of vicarious liability principles); Civil Code of FRANCE, Art. 1384 (1994) (“[A person] is liable not only for the damage which he caused by his own act, but also for that which is caused by the act of persons for whom he is responsible, or by things which he has in his keeping.”); Civil Code op GeRmany, § 831 (1975) (“A person who employs another to do any work is bound to compensate for any damage which the other unlawfully causes to a third party in the performance of his work.”). Thus, the conclusion that plaintiffs’ agency theory is cognizable under federal common law is further supported by the international legal authorities that establish agency as a general principle of international law.
A factual question requiring trial exists with respect to whether an agency relationship existed between Unocal and the Myanmar military. Some evidence in the record suggests that such a relationship existed. For instance, plaintiffs cite an internal Unocal briefing document regarding the Yadana Project, discussed by the majority, which states that”[a]ccording to our contract, the government of Myanmar is responsible for protecting the pipeline. There is military protection for the pipeline and, when we have work to do along the pipeline that requires security, the military people will, as a matter of course, be nearby.” They also point to memoranda by various Total and Unocal employees recounting that oil company officials requested specific battalions to perform various tasks, including the construction of helipads for the convenience of corporate executives. Plaintiffs argue that the record supports either an implied or express agency relationship, based on the conduct of the parties. As the majority has also pointed out, the record contains evidence of daily meetings between Total and Unocal executives and Myanmar military commanders, so that the corporations could instruct the military leaders regarding specific security or infrastructure projects that were required for the pipeline construction. Moreover, Unocal stated publicly on several occasions that it controlled the Myanmar military’s actions in connection with the pipeline project. In response to accusations of human rights abuses occurring by the Myanmar military with respect to the Yadana project, Unocal denied the existence of such abuses, and stressed its ability to prevent any wrongdoing due to its control of the military. Unocal’s alleged actions directing the Myanmar military create a triable question of fact as to whether an agency relationship existed between Unocal and the Myanmar armed forces.
*974Significantly, there is evidence in the record that Unocal did not simply, by its inaction, take advantage of the routine presence of the Myanmar military in the Tenasserim region. The record indicates that there was in fact little to no rebel activity in the region where the pipeline construction occurred, and that the center of the Myanmar civil war was 150-200 miles distant from the pipeline project. A jury could reasonably conclude that the military was present in Tenasserim not merely to maintain order, as was its function in other parts of the nation. Rather, it could determine that military forces were brought to Tenasserim in order to support the pipeline project, that the military was performing duties for the pipeline project quite distinct from traditional military or state functions, and that it did so at the request of and in close coordination with Unocal and the other private entities. It is not essential that a formal contract have existed between Unocal and the Myanmar military in order for Unocal to be held liable for the government’s actions under an agency theory. Nevertheless, should plaintiffs prove their allegation that such a contract existed, a -jury might have considerable difficulty in accepting Unocal’s denial of an agency relationship.
c. Reckless Disregard
Finally, the facts alleged by the plaintiffs, if proved, support a recovery against Unocal under an additional theory, that of the common law theory of recklessness or reckless disregard. Here, plaintiffs allege that Unocal had actual knowledge that the Myanmar military would likely engage in human rights abuses, including forced labor, if it undertook the functions Unocal and the other private parties desired it to perform in connection with the Yadana Pipeline Project. Nevertheless, according to plaintiffs, Unocal recklessly disregarded that known risk, determined to use and in fact did use the services of that military to perform pipeline-related tasks, and thereby set in motion international law abuses that were foreseeable to Unocal. Plaintiffs thus allege that Unocal acted with recklessness, which occurs when a party is aware of (or should be aware of) an unreasonable risk, yet disregards it, thereby leading to harm to another. Farmer v. Brennan, 511 U.S. 825, 836, 114 S.Ct. 1970,128 L.Ed.2d 811 (1994).
Federal common law contains two variants of the theory of recklessness or reckless disregard. Plaintiffs include both in their complaint, and in my view the record contains evidence sufficient to require trial on both. The first is traditional civil-law recklessness, sometimes referred to as “objective recklessness”; the Supreme Court has stated that”[t]he civil law generally calls a person reckless who acts or (if the person has a duty to act) fails to act in the face of an unjustifiably high risk of harm that is either known or so obvious that it should be known.’’ Farmer, 511 U.S. at 836-37, 114 S.Ct. 1970 (citing W. Keeton, D. Dobbs, R. Keeton, & D. Owen, PROSSER AND KEETON ON LAW OF TORTS § 34, pp. 213-214 (5th ed.1984); Restatement (Second) of Torts § 500 (1965)). The second version is “subjective recklessness,” also referred to as “willful recklessness.” This doctrine requires actual knowledge of a substantial risk which the defendant subsequently disregards. Ewolski v. City of Brunswick, 287 F.3d 492, 513 (6th Cir.2002).
The concept that one party may be held liable for a reckless disregard of the welfare of another pervades federal common law and has been applied in a variety of contexts. It is recognized, for instance, in admiralty law cases, see, e.g., Youell v. Exxon Corp., 48 F.3d 105, 110-11 (2d Cir.), vacated on other grounds, 516 U.S. 801, 116 S.Ct. 43, 133 L.Ed.2d 9 (1995), in cases arising under the Warsaw Convention, Saba v. Compagnie Nationale Air France, *97578 F.3d 664, 668-69 (D.C.Cir.1996), as well as in constitutional tort cases brought pursuant to 42 U.S.C. § 1983. Medina v. City and County of Denver, 960 F.2d 1493, 1496 (10th Cir.1992). The Supreme Court has also adopted a willful recklessness standard in cases involving “deliberate indifference” under the Eighth Amendment. Farmer, 511 U.S. at 836, 114 S.Ct. 1970 (“With deliberate indifference lying somewhere between the poles of negligence at one end and purpose or knowledge at the other, the Courts of Appeals have routinely equated deliberate indifference with recklessness. It is, indeed, fair to say that acting or failing to act with deliberate indifference to a substantial risk of serious harm to a prisoner is the equivalent of recklessly disregarding that risk.” (citations omitted)). On “a continuum that runs from simple negligence through gross negligence to intentional misconduct,” Saba, 78 F.3d at 668, recklessness lies between gross negligence and intentional harm. The common law principle of recklessness has typically been applied to acts by a defendant that directly cause harm to a plaintiff. Nevertheless, I see no reason why the general principle that liability arises for one party’s conscious disregard of unreasonable risks to another should not apply when a defendant consciously disregards the risks that arise from its decision to use the services of an entity that it knows or ought to know is likely to cause harm to another party.13
Proof of even willful recklessness does not require proof of intent; it requires only that a defendant have acted in conscious disregard of known dangers. City of Canton, Ohio v. Harris, 489 U.S. 378, 390 n. 10, 109 S.Ct. 1197, 103 L.Ed.2d 412 (1989). The doctrine of reckless disregard of another’s rights has been well-developed in the § 1983 context; there, courts have held that a plaintiff need not prove that a defendant intended to cause harm to the specific plaintiff. “[R]eckless intent does not require that the actor intended to harm a particular individual; reckless intent is established if the actor was aware of a known or obvious risk that was so great that it was highly probable that serious harm would follow and he or she proceeded in conscious and unreasonable disregard of the consequences. Thus, reckless intent involves disregard of a particular risk rather than intent to cause a particularized harm.” Medina, 960 F.2d at 1496. Plaintiffs’ theory that by using the services of the Myanmar military in connection with the Yadana Project, Unocal recklessly disregarded the likelihood that their human rights would be violated is thus well grounded in federal common law.
Plaintiffs have presented sufficient evidence to proceed to trial on the reckless disregard claim. There exists a genuine question of material fact as to whether the Myanmar military caused the human rights abuses alleged, and if so, whether Unocal should be liable for those abuses because it acted in either a subjectively or an objectively reckless manner in choosing to make use of the services of the Myanmar military in connection with the Yadana gas exploration project. Plaintiffs *976allege that Unocal entered into an agreement to participate in the Yadana project knowing that the Myanmar military was also a participant, and that the Myanmar military would perform important pipeline tasks for-the project, even though it had knowledge that the military engaged in widespread human rights abuses, including forced labor. Plaintiffs further allege that Unocal continued to use those services after obtaining knowledge that the military was conscripting forced labor in service of the Yadana project.
The record contains evidence that Unocal held several meetings with human rights groups both before and after it became a part of the Yadana venture, at which those organizations provided Unocal with documentation of the Myanmar military’s brutal treatment of the Myanmar people. Moreover, as noted earlier, Unocal retained a risk management consulting firm prior to joining the Yadana project, and that firm completed a report informing the company that “the government habitually makes use of forced labor,” and assessing the risk that the proposed joint venture would result in the use of forced labor as “high.”
Additionally, as the majority has noted, Unocal .Vice President Lipman testified that even before Unocal invested in the Project, it was aware that “the option of having the [Myanmar] [M]ilitary provide protection for the pipeline construction and operation of it would be that they might proceed in the manner that would be out of our control and not be in a manner that we would like to see them proceed, I mean, going to- excess.” (Emphasis added.) Unocal Representative Robinson later wrote to Unocal President Imle that “[o]ur assertions that [the Myanmar Military] has not expanded and amplified its usual methods around the pipeline on our behalf may not withstand much scrutiny.” (Emphasis added.) And Unocal Consultant Haseman told Unocal that “[t]he most common -[human rights violations] are forced relocation without compensation of families from land near/along the pipeline route; forced labor to work on infrastructure projects supporting the pipeline ...; and ... execution by the army of those opposing such actions.” Based on this evidence, a question , of fact exists with respect to whether consequences that were or should have been foreseeable to Unocal, in the form of violations of international law by the Myanmar military, resulted from Unocal’s participation in the Yadana Pipeline Project, and from its continuing decision to use the services of the Myanmar military during the course of the pipeline’s construction.14
d. Murder and Rape Claims15
Like the majority, I agree with the Second Circuit’s holding in Kadic v-. Karadzic, *977that under the Alien Tort Claims Act, a plaintiff may recover for wrongs that occur ancillary to a violation of international human rights law as part of the claim for the primary violation. 70 F.3d 232, 244 (2d Cir.1995) (holding acts of rape, torture, and summary execution committed in connection with genocide or war crimes to be actionable under the ATCA). Here, all of the acts of murder and rape of plaintiffs alleged in the complaint or otherwise in the record on summary judgment occurred in furtherance of the forced labor program. Following the Kadic rule, plaintiffs may pursue those claims as part of their forced labor claims.
As with the forced labor claims, however, I disagree with the majority regarding what it is necessary for plaintiffs to prove in order for Unocal to be held liable for acts of murder or rape. Specifically, I disagree with the majority’s view that we must apply once again, independently, a third-party liability standard — whether international law or federal common law— this time with respect to the specific acts of murder and rape. In my opinion, if it is established that the alleged rapes and murder of plaintiffs occurred in furtherance of the forced labor program, and if Unocal is held hable for the forced labor practices of the Myanmar military, then plaintiffs need not again prove separately the elements of a third-party liability theory. In such case, they need prove only the additional facts supporting the rape and murder allegations. While I would not foreclose a possible foreseeability or ultra vires argument, I would not, as the majority does, require plaintiffs to make a second showing of third-party liability merely because specific acts conducted in furtherance of the primary tort are themselves tortious.
There is one final observation that I hope will clarify what the majority opinion does and does not do. Because the acts of murder and rape involving the plaintiffs all allegedly occurred in furtherance of a regime of forced labor, there is no need to address the question whether Unocal could be held liable if members of the Myanmar military had committed similar acts against plaintiffs separate and apart from a forced labor regime or a similarly well-established pattern of conduct violative of international law. Cf Kadic, 70 F.3d at 244-45 (leaving open the related question of whether non-state actors who directly commit acts such as murder, rape, or torture may be held liable under the ATCA when those acts are not committed in furtherance of a jus cogens violation.). Accordingly, I read the majority’s statements regarding the rapes and murders as leaving open the question whether a private entity could be held liable for such government conduct if it was unrelated to an underlying violation of international law.
A Conclusion
In sum, I agree with the majority that disputed questions of fact exist with respect to whether human rights violations occurred during the construction of the Yadana Pipeline Project, and with respect to the nature of Unocal’s involvement in such violations. Assuming the necessary evidence is introduced at trial, I would, however, direct that the jury be instructed to apply the three common law theories of third-party liability ordinarily applied in *978tort cases, rather than the international criminal law doctrine of aiding and abetting, including by means of moral support, recently announced by the ad hoc War Crimes Tribunal for the Former Yugoslavia.
. As the majority notes, such a norm is one that is “accepted and recognized by the international community of states as a whole as a norm from which no derogation is permitted and which can be modified only by a subsequent norm of general international law having the same character.” Vienna Convention on the Law of Treaties, art. 53, May 23, 1969, 1155 U.N.T.S. 332, 8 I.L.M. 679; see also Siderman de Blake v. Republic of Argentina, 965 F.2d 699, 714-15 (9th Cir.1992)
. I do not read the majority opinion as holding otherwise. The opinion nowhere states that a third party can be liable for a governmental action only if that action constitutes a jus cogens violation. Rather, I view the majority’s interesting academic dissertation on why forced labor is like slavery and why slavery constitutes a jus cogens violation as background historical material that may be of interest to legal scholars but is unnecessary to its decision.
. It is the rare Alien Tort Claims Act case that does not involve a foreign state or official as a defendant. Most international law norms apply only to states; a private party will ordinarily violate international law by its own actions only if it transgresses a legal norm that has achieved jus cogens status. See discussion, supra.
. That_the principles discussed in the text " apply not only to traditional domestic legislation but to the Alien Tort Claims Act as well is demonstrated by a statement by the Eleventh Circuit in Abebe-Jira v. Negewo, 72 F.3d 844, 848 (11th Cir.1996). There, that court observed that the purpose of the ATCA is "to establish! 1 a federal forum where courts may fashion domestic common law remedies to give effect to violations of customary international law.”
. The International Criminal Tribunal for the Former Yugoslavia was formed with the limited mandate of adjudicating allegations of human rights abuses that took place in the Balkans in the last decade. Established by Security Council Resolution 827 in May, 1993, it is a temporary body whose members are elected for four-year terms by the members of the United Nations General Assembly. The International Criminal Tribunal for Rwanda, which subsequently applied the Yugoslav Tribunal’s test, is a similarly-constituted body.
. The Restatement (Second) of Conflict of Laws, § 6, provides that the following factors are to be considered as part of a choice of law analysis: (1) the needs of the interstate and international systems; (2) the relevant policies of the forum, (3) the relevant policies of other interested states and the relative interests of those states in the determination of the particular issue; (4) the protection of justified expectations, (5) the basic policies underlying the particular field of law, (6) certainty, predictability and uniformity of result, and (7) ease in the determination, and application of the law to be applied. “Federal choice of law rules follow the approach of the Restatement (Second) of Conflict of Laws.’’ In re Vortex Pishing Sys., Inc., 277 F.3d 1057, 1069 (9th Cir.2002); see also Chuidian v. Philippine Nat’l. Bank, 976 F.2d 561, 564 (9th Cir.1992)
. The two remaining Restatement factors are the first, “the needs of the interstate and international systems,” and the third, "the relevant policies of other interested states and the relative interests of those states in the determination of the particular issue.” Restatement (Second) of Conflict of Laws, § 6.
. As the Second Circuit explained in the landmark case of Filartiga, all international legal principles do not automatically become a part of the federal common law; only those that achieve the status of customary international law or are included in international treaties are incorporated as part of federal common law. A customary international law rule "results from a general and consistent practice of states followed -by them from a sense of legal obligation.” Restatement (Third) Foreign Relations Law, § 102. The Filartiga court observed, “[t]he requirement that a rule command the 'general assent of civilized nations' to become binding upon them all is a stringent one. Were this not so, the courts of one nation might feel free to impose idiosyncratic legal rules upon others, in the name of applying international law.” 630 F.2d at 881 (quoting The Paquete Habana, 175 U.S. at 694, 20 S.Ct. 290). Thus, the Filartiga court reached its conclusion that a claim for torture was cognizable under the ATCA in large part because the prohibition on torture had - become part of customary international law. Id. ■ -
.In an effort to minimize the damage caused by its unfortunate decision to apply international law to the third-party liability issue, and perhaps to make that choice of law more palatable to American courts generally, the majority disclaims an integral portion of the international law standard it adopts, purporting to leave “to another day” the question whether moral support alone (whatever that may mean) is sufficient to give rise to third-party liability. See maj. op., 951. However, by substituting international law standards for federal common law, rather than following federal common law and incorporating those portions of international law that attract sufficient legal support, the majority has lost whatever opportunity it had to pick and choose the aspects of international law that it finds appealing. Having declared that international law governs, and that the Yugoslav Tribunal's standard constitutes the controlling international law, the majority cannot then escape the implications of being bound by the law it has selected. Indeed, the majority, despite its disclaimer as to "moral support,” is no more successful in avoiding the consequences of its choice of the Yugoslav Tribunal's criminal standard than it is in avoiding the consequences of its predicate decision to reject federal common law as the appropriate rule for ancillary issues in Alien Tort Claims Act cases. See maj. op., n. 25.
. For instance, liability for moral support raises the question whether political advocacy not imminently causing violence that would otherwise be protected by the First Amendment could be the source of ATCA liability under the majority’s standard.
. It is well-accepted that joint liability will exist where (1) parties intended to form a joint venture; (2) parties share a common interest in the subject matter of the venture; (3) the parties share the profits and losses of the venture; and (4) the parties have joint *971control or the joint right of control over the venture. W. Keeton, Prosser & Keeton on Torts, § 72 at 518 (5th ed.1984).
. In addition, the Foreign Sovereign Immunities Act, 28 U.S.C. § 1605, demonstrates that agency liability is a general principle applicable to international entities. In that statute, an exception exists to the general sovereign immunity accorded to foreign states in American courts for certain commercial activities of agents of foreign nations. Id. § 1605(a)(3).
. I reach this conclusion in part because at common law, a defendant may be liable for harms caused by an entity that it negligently employs, even if no respondeat superior or agency relationship exists. See, e.g., Bennett v. United States, 803 F.2d 1502, 1505 (9th Cir.1986) (holding that the U.S. government may be liable for damages arising from the kidnapping and raping of several children by a teacher hired by the government where the government knew or should have known that the teacher had a history of child molestation); see also Van Ort v. Estate of Stanewich, 92 F.3d 831, 837 (9th Cir.1996) (stating that under Monell v. Department of Social Services, 436 U.S. 658, 98 S.Ct. 2018, 56 L.Ed.2d 611 (1978), a municipality may be liable for the negligent hiring of municipal employees).''
. The district court granted summary judgment for defendants in part because in its view plaintiffs produced insufficient evidence that Unocal proximately caused plaintiffs' injuries. However, under none of the three federal common law theories discussed in the text is proximate cause a necessary element of Unocal's liability. If proximate cause applies at all, it applies to the question whether the Myanmar military was responsible for the international law violations. As a practical matter, if the alleged violations occurred there can be little doubt that they were proximately caused by the Myanmar military. As explained in the text, under any of the three theories, Unocal’s liability is indirect; proximate cause is therefore irrelevant to plaintiffs’ claims against Unocal.
. I agree with the majority’s conclusion that there is insufficient evidence in the record to permit plaintiffs to proceed on a claim of torture, and would limit their claims to those alleging forced labor, murder, and rape. I would also note that the record is replete with horrific accounts of physical abuse of Myanmar villagers by members of the military. Because the victims in those accounts are not plaintiffs in this action, and this suit does not *977constitute a class action, many substantial allegations of wrongdoing may not give rise to liability as a part of this case.