FOR PUBLICATION

UNITED STATES COURT OF APPEALS

FOR THE NINTH CIRCUIT

TAHAWWUR HUSSAIN RANA, No. 23-1827 D.C. No. Petitioner - Appellant, 2:23-cv-04223-

DSF

v.

W.Z. JENKINS II, OPINION

Respondent - Appellee.

Appeal from the United States District Court for the Central District of California Dale S. Fischer, District Judge, Presiding

Argued and Submitted June 5, 2024 Pasadena, California

Filed August 15, 2024

Before: MILAN D. SMITH, JR. and BRIDGET S. BADE, Circuit Judges, and SIDNEY A. FITZWATER, District Judge. *

Opinion by Judge Milan D. Smith, Jr.

* The Honorable Sidney A. Fitzwater, United States District Judge for the Northern District of Texas, sitting by designation. SUMMARY **

Habeas Corpus

The panel affirmed the district court’s denial of Tahawwur Hussain Rana’s 28 U.S.C. § 2241 habeas corpus petition challenging a magistrate judge’s certification of Rana as extraditable to India for his alleged participation in terrorist attacks in Mumbai. Under the limited scope of habeas review of an extradition order, the panel held that Rana’s alleged offense fell within the terms of the extradition treaty between the United States and India, which included a Non Bis in Idem (double jeopardy) exception to extraditability “when the person sought has been convicted or acquitted in the Requested State for the offense for which extradition is requested.” Relying on the plain text of the treaty, the State Department’s technical analysis, and persuasive case law of other circuits, the panel held that the word “offense” refers to a charged crime, rather than underlying acts, and requires an analysis of the elements of each crime. The panel concluded that a coconspirator’s plea agreement did not compel a different result. The panel held that the Non Bis in Idem exception did not apply because the Indian charges contained distinct elements from the crimes for which Rana was acquitted in the United States. The panel also held that India provided sufficient competent evidence to support the magistrate judge’s

** This summary constitutes no part of the opinion of the court. It has been prepared by court staff for the convenience of the reader.

RANA V. JENKINS 3

finding of probable cause that Rana committed the charged crimes.

COUNSEL

John D. Cline (argued), Law Office of John D. Cline, Seattle, Washington; Jennifer L. Williams, Summa LLP, South Pasadena, California; for Petitioner-Appellant. Bram M. Alden (argued), Assistant United States Attorney Chief, Criminal Appeals Section; David R. Friedman and John J. Lulejian, Assistant United States Attorneys; E. Martin Estrada, United States Attorney; Department of Justice, Office of the United States Attorney, Los Angeles, California; Kerry A. Monaco, Trial Attorney; Rebecca A. Haciski, Associate Director; Office of International Affairs; Bruce C. Swartz, Deputy Assistant Attorney General; Nicole M. Argentieri, Acting Assistant Attorney General, Criminal Division; United States Department of Justice, Washington, D.C.; for Respondent-Appellee.

OPINION

M. SMITH, Circuit Judge:

Tahawwur Hussain Rana, a Pakistani national, was tried in a United States district court on charges related to his support for a terrorist organization that carried out large- scale terrorist attacks in Mumbai, India. A jury convicted Rana of providing material support to a foreign terrorist organization and conspiring to provide material support to a foiled plot to carry out terrorist attacks in Denmark. However, the jury acquitted Rana of conspiring to provide material support to terrorism related to the attacks in India. After Rana served seven years in prison for those convictions and upon his compassionate release, India issued a request for his extradition to try him for his alleged participation in the Mumbai attacks. Before the magistrate judge who initially decided Rana’s extraditability (the extradition court), Rana argued that the United States’ extradition treaty with India protected him from extradition because of its Non Bis in Idem (double jeopardy) provision. He also argued that India did not provide sufficient evidence to demonstrate probable cause that he committed the charged crimes. The extradition court rejected Rana’s arguments and certified that he was extraditable. After Rana raised the same arguments in a habeas petition in district court (the habeas court), the habeas court affirmed the extradition court’s findings of facts and conclusions of law. This appeal timely followed, with Rana raising both his Non Bis in Idem and probable cause arguments. We affirm.

FACTUAL AND PROCEDURAL BACKGROUND

I. Factual Background

Rana and David Coleman Headley were childhood friends. 1 In adulthood, Rana deserted his post in the Pakistani army and moved to Chicago, where he established an immigration business. For his part, Headley started trafficking heroin, ultimately radicalized, and attended

1 We recount these facts based on testimony and evidence before the extradition court, including testimony and evidence presented in Rana’s criminal trial in the Northern District of Illinois. Rana disputes their accuracy, as outlined in Section II below.

RANA V. JENKINS 5

training camps operated by the terrorist organization Lashkar-e-Tayyiba (Lashkar). According to Headley’s testimony, the pair met in Chicago multiple times over the course of three years, plotting to assist Lashkar in terrorist attacks that ultimately killed and injured hundreds of people. In August 2005, the pair met over several days in Chicago, where Headley told Rana about Lashkar’s plans for Headley to travel to public places and government facilities in India to conduct surveillance for a possible attack. Headley proposed using Rana’s immigration business as a front for Lashkar’s surveillance activities, with Headley posing as an “immigration consultant” for Rana in Mumbai. To sweeten the deal for Rana, Headley offered to help resolve Rana’s status as a deserter from the Pakistani army. In June 2006, Headley again met with Rana in Chicago, elaborating on his involvement with Lashkar. After agreeing to open a Mumbai branch of his immigration business, Rana helped Headley complete a successful application for an Indian business visa, which contained several inaccuracies. Headley used the visa to travel to India under the pretense of operating the Mumbai branch of Rana’s business. Although Headley rented an apartment, hired a secretary, and signed a lease for the branch, little to no immigration work occurred there. In July 2007, Headley stayed at Rana’s Chicago home, informed him about the surveillance he had conducted while in India, and showed Rana a video Headley had taken of the Taj Mahal Palace Hotel. Rana helped Headley secure a five- year multi-entry Indian visa. Utilizing that visa, Headley traveled to India multiple times between September 2007 and March 2008, conducting further surveillance of potential targets. In May 2008, Headley informed Rana about the surveillance he conducted in Mumbai, identifying possible attack targets, including the Taj Mahal Palace Hotel. In the fall of 2008, Headley warned Rana to avoid India, where attacks were imminent, and arranged for Rana to meet with one of their co-conspirators in Dubai. During a later intercepted conversation, Rana told Headley that their co- conspirator in Dubai had confirmed the upcoming attacks. The lease on the Mumbai office expired in November 2008, and neither Rana nor Headley renewed it. Lashkar carried out massive terrorist attacks throughout Mumbai, including at the Taj Mahal Palace Hotel, between November 26 and 29, 2008. The attacks killed 166 people, injured 239, and resulted in more than $1.5 billion in property damage. Rana commended the terrorists who carried out the attacks and stated that the people of India “deserved it.” After the terrorist attacks in India were completed, Headley and Lashkar began plotting new, but ultimately unsuccessful, attacks in Denmark and India. Headley once again used the immigration business as a cover to conduct surveillance, this time in Denmark. Headley kept Rana apprised of his surveillance activities, and Rana communicated directly with Headley’s Lashkar contacts.

II. Procedural Background

On October 3, 2009, United States law enforcement arrested Headley in Chicago. Headley pled guilty to twelve terrorism-related charges in the Northern District of Illinois, including multiple counts related to the Mumbai attacks and the foiled Denmark plot. Headley agreed to cooperate with the United States, and his plea agreement contained a non- extradition provision.

RANA V. JENKINS 7

Law enforcement also arrested and indicted Rana on October 18, 2009. Rana was tried on three counts: conspiracy to provide material support to terrorism in India, 18 U.S.C. § 2339A, conspiracy to provide material support to terrorism in Denmark, id., and providing material support to a foreign terrorist organization, id. § 2339B. Headley testified for the prosecution. On June 9, 2011, the jury convicted Rana of the terrorism conspiracy related to Denmark and providing material support to Lashkar but acquitted him of the terrorism conspiracy related to India. On January 17, 2013, the district court sentenced Rana to 14 years in prison. After serving seven years of prison time, Rana’s motion for compassionate release was granted during the COVID-19 pandemic. While Rana was in United States custody, the Indian government charged Rana, accusing him of conspiring to plan and carry out the Mumbai attacks. On August 28, 2018, an Indian judge issued a warrant for Rana’s arrest on charges related to the attacks, including (1) conspiracy to (a) wage war, (b) commit murder, (c) commit forgery for the purpose of cheating, (d) use as genuine a forged document or electronic record, and (e) commit a terrorist attack; (2) waging war; (3) murder; and (4) committing a terrorist act. 2 India subsequently requested Rana’s extradition. One day after Rana was granted compassionate release, the United States filed a complaint for his provisional arrest in response to India’s extradition request. On May 16, 2023,

2 We have simplified the charges for the purposes of this opinion. Some of the charges, including conspiracy to wage war and conspiracy to commit a terrorist act, were brought as multiple charges under separate provisions of India’s Penal Code and Unlawful Activities Prevention Act. the extradition court certified Rana’s extradition pursuant to 18 U.S.C. § 3184 and rejected Rana’s claims that (1) his extradition to India was barred under the Non Bis in Idem provision of the Extradition Treaty Between the Government of the United States of America and the Government of the Republic of India (the Treaty) and (2) India’s evidence against Rana failed to establish probable cause that Rana committed the offenses for which the certification of extradition was sought. Rana sought collateral review in the habeas court by filing a petition pursuant to 28 U.S.C. § 2241. The habeas court rejected Rana’s same arguments, denying his petition on August 10, 2023. The habeas court, however, stayed Rana’s extradition pending this timely appeal.

JURISDICTION AND STANDARD OF REVIEW

The extradition court had jurisdiction pursuant to 18 U.S.C. § 3184. The habeas court had jurisdiction pursuant to 28 U.S.C. § 2241(a). We have jurisdiction pursuant to 28 U.S.C. §§ 1291, 2253(a). “We review de novo the district court’s denial of a habeas petition in extradition proceedings.” United States v. Knotek, 925 F.3d 1118, 1124 (9th Cir. 2019). We evaluate “factual questions, as determined by the extradition magistrate judge, for clear error.” McKnight v. Torres, 563 F.3d 890, 892 (9th Cir. 2009). “The scope of habeas review of an extradition order is severely limited.” Artukovic v. Rison, 784 F.2d 1354, 1355– 56 (9th Cir. 1986). We can review only “whether: (1) the extradition magistrate judge had jurisdiction over the individual sought, (2) the treaty was in force and the accused’s alleged offense fell within the treaty’s terms, and

RANA V. JENKINS 9

(3) there is any competent evidence supporting the probable cause determination of the magistrate judge.” Knotek, 925 F.3d at 1124 (citation, internal quotation marks, and alterations omitted). Rana challenges only the latter two issues, and we address them in turn.

ANALYSIS

I. Rana’s alleged offense fell within the Treaty’s terms.

Article 1 of the Treaty states that the United States and India “agree to extradite to each other . . . persons who, by the authorities in the Requesting State are formally accused of, charged with or convicted of an extraditable offense . . . .” One exception is “when the person sought has been convicted or acquitted in the Requested State for the offense for which extradition is requested,” codified at Article 6(1). Rana argues that he cannot be extradited based on conduct for which he was acquitted in the United States because the word “offense” refers to underlying acts. The government argues that “offense” refers to a charged crime and requires an analysis of the elements of each charged crime. Thus, according to the government, the Treaty permits Rana’s extradition because the Indian charges contain distinct elements from the crimes for which he was acquitted in the United States. Treaty interpretation begins with the Treaty’s plain text, but we also consider its negotiations, drafting history, and post-ratification understanding of the signatory nations. Medellín v. Texas, 552 U.S. 491, 506–07 (2008). Here, the Treaty’s plain terms, the post-ratification understanding of the signatories, and persuasive precedent all support the government’s interpretation. Rana argues, however, that, based on the government’s interpretation of the Treaty in Headley’s plea agreement, we should judicially estop the government from advocating for its current interpretation of the Treaty. We decline to do so. A. The plain text of the Treaty supports a meaning of “offense” that denotes a charged crime defined by its elements. The parties refer us to several provisions both within and outside the Treaty that assist us in interpreting Article 6(1): Article 6(2), Article 2, and the United States’ other extradition treaties. We address each source in turn. First, Article 6, when read as a whole, compels a reading of “offense” that requires comparing the elements of each country’s crimes. We begin with the text. 3 Article 6 contains two provisions. Paragraph 1, the provision we are tasked with interpreting, states in full: “Extradition shall not be granted when the person sought has been convicted or acquitted in the Requested State for the offense for which extradition is requested.” (emphasis added). The subsequent provision, Paragraph 2, states: “Extradition shall not be precluded by the fact that the authorities in the Requested State have decided not to prosecute the person sought for the acts for which extradition is requested . . . .” (emphasis added). As is apparent, Paragraph 1 uses the word “offense,” while the very next provision uses the word “acts.” When treaties use differing language in parallel provisions, it “implies that the drafters of the [Treaty] understood the word[s] [] to mean something different . . . for they otherwise logically would have used the same word in each article.” Air France v. Saks, 470 U.S. 392, 398 (1985);

3 The government points to Black’s Law Dictionary, which defines “offense” as a “violation of the law” or “crime.” Black’s Law Dictionary 1300 (11th ed. 2019). This definition does not aid our inquiry because it begs the question of how we should compare similar crimes.

RANA V. JENKINS 11

Hosaka v. United Airlines, Inc., 305 F.3d 989, 995 (9th Cir. 2002). Thus, the most natural reading of Paragraph 1 compels a definition of “offense” that is distinct from “acts.” Because “acts” in Paragraph 2 refers to uncharged conduct, “offense” in Paragraph 1 must refer to something other than uncharged conduct. Rana argues that “acts” in Paragraph 2 means the same thing as “offense[s]” as used in Paragraph 1. This argument is unpersuasive. As Rana concedes in his reply brief, “Article 6(2) refers to instances where the requested state has declined to prosecute, meaning that the acts at issue may never have coalesced into an offense under the laws of that state.” We thus read “offense” as a charged crime, with elements, as distinct from uncharged “acts” or conduct. 4 Second, Rana points to Article 2’s dual criminality provision, which uses “offense” to refer to underlying conduct. Article 2 states that “[a]n offense shall be an extraditable offense if it is punishable under the laws in both Contracting States by deprivation of liberty, including imprisonment, for a period of more than one year or by a more severe penalty.” Binding Ninth Circuit precedent instructs that, for purposes of the dual criminality provision, “the elements of the analogous offenses need not be identical.” Clarey v. Gregg, 138 F.3d 764, 766 (9th Cir. 1998). Rana argues that because “offense” is used to refer

4 Rana’s argument that similar treaties’ submittal letters refer to “acts” as “offenses” when explaining similar provisions does not cut in his favor. While it is possible (although unlikely) that “acts” and “offenses” are intended as synonyms in Article 6, that reading does not necessarily mean that they both refer to “conduct.” It is equally as likely that both words refer to crimes consisting of elements. to conduct in Article 2, it must also refer to conduct in Article 6. While Rana is correct that typically “identical words used in different parts of the same act are intended to have the same meaning,” Sorenson v. Secretary of the Treasury, 475 U.S. 851, 860 (1986), Article 2 contains limiting language. It instructs that “offense” is to be interpreted “[f]or the purposes of this Article” “whether or not the laws in the Contracting States place the offense within the same category of offenses or describe the offense by the same terminology.” In other words, the Treaty explicitly states that for the purposes of the dual criminality provision, whether the offenses in each state use the same terminology (or, in other words, elements) is irrelevant. See Knotek, 925 F.3d at 1131 & n.12 (explaining that a similar provision in the extradition treaty with the Czech Republic incorporated the rule that, for the purposes of dual criminality, “[t]he elements of one offense ‘need not be identical to the elements of a similar offense in the United States’”). No such limiting language exists in Article 6. Moreover, despite Rana’s argument that “offense” consistently means “acts” in Article 2, multiple provisions seem to reference the elements of a crime when using “offense.” For example, Article 2, Paragraph 4 of the Treaty states: “Extradition shall be granted for an extraditable offense regardless of where the act or acts constituting the offense were committed.” The Treaty here acknowledges that various acts make up an offense. The contracting states could have drafted language requiring that extradition shall be granted for an extraditable offense “regardless of where the offense was committed,” but they did not. Further, Paragraph 2 of Article 2 states: “An offense shall also be an extraditable offense if it consists of an attempt or a

RANA V. JENKINS 13

conspiracy to commit, aiding or abetting, counselling or procuring the commission of or being an accessory before or after the fact to, any offense described in paragraph 1.” Here, Article 2, Paragraph 2 refers to an offense as something that includes elements, such as aiding and abetting or conspiracy. Third, this interpretation is further supported by comparing treaties that specifically use the word “acts” in their Non Bis in Idem provisions. For example, the U.S. Extradition Treaty with Italy uses the word “acts” in its Non Bis in Idem provision, as did the 1971 Extradition Treaty between the United States and France. When the United States and France rewrote their extradition treaty in 1996, they changed the Non Bis in Idem provision to use the term “offense.” These examples demonstrate that, when the United States and its allies want to protect similar conduct from being dually prosecuted in each nation, they use language to evince that intent. They did not do so here, and we see no reason to ignore the Treaty’s plain terms. B. The State Department’s technical analysis supports an elements approach. The post-ratification understanding of the signatories confirms this reading. “It is well settled that the Executive Branch’s interpretation of a treaty ‘is entitled to great weight.’” Abbott v. Abbott, 560 U.S. 1, 15 (2010) (quoting Sumitomo Shoji America, Inc. v. Avagliano, 457 U.S. 176, 185 (1982)). Here, the United States Departments of State and Justice submitted a technical analysis to Congress based on their notes in negotiating the Treaty. The analysis states that Article 6, Paragraph 1 “applies only when the person sought has been convicted or acquitted in the Requested State of exactly the same crime that is charged in the Requesting State. It is not enough that the same facts were involved.” (emphasis added). We give this interpretation substantial weight. 5 Prior to the Supreme Court's overruling of Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984), see Loper Bright Enters. v. Raimondo, 144 S. Ct. 2244 (2024), Rana argued that Chevron’s likely impending demise would undermine the precedent requiring us to defer to the Executive Branch’s interpretation of the Treaty. But because the logic underpinning Chevron deference is entirely distinct from the logic underpinning a deference to the Executive in matters of foreign affairs, see U.S. Const. art. II, § 2; Haig v. Agee, 453 U.S. 280, 293–94 (1981) (“[T]he generally accepted view [is] that foreign policy was the province and responsibility of the Executive.”), Loper Bright has no effect on our decision here. Documents like a technical analysis help us understand the negotiators’ intent in creating a binding agreement with another nation. We defer to those documents because they state what the drafters meant when they wrote the treaties at issue. As we explained in Patterson v. Wagner, 785 F.3d 1277 (9th Cir. 2015):

Because the purpose of treaty interpretation is to ‘give the specific words of the treaty a meaning consistent with the shared expectations of the contracting parties,

5 This interpretation is further supported by India’s similar reading of the Treaty. The Indian opinion that Rana claims the government overlooks is a dissenting opinion from the Indian Supreme Court and lacks the force of law.

RANA V. JENKINS 15

courts—including our Supreme Court—look to the executive branch’s interpretation of the issue, the views of other contracting states, and the treaty’s negotiation and drafting history in order to ensure that their interpretation of the text is not contradicted by other evidence of intent.

Id. at 1281–82 (internal citations omitted); see, e.g., Manrique v. Kolc, 65 F.4th 1037, 1043 (9th Cir. 2023) (utilizing technical analysis as “strong evidence” of the United States’ intent). On the other hand, Chevron deference relied primarily on an agency’s policy expertise, rather than its insight into Congress’s intent. See Chevron, 467 U.S. at 865–66 (explaining that judges, who are not policy experts, should defer to an agency’s wise policy expertise). Loper Bright held that this long-standing rationale was flawed because the Administrative Procedure Act’s (APA) mandate that reviewing courts determine “all relevant questions of law” means that courts must “exercise independent judgment in determining the meaning of statutory provisions.” 144 S. Ct. at 2262. This reasoning does not touch, let alone undermine, the principle that we are to give deference to the Executive Branch’s understanding of its own treaties. Rana makes similar arguments based on his view that the Supreme Court is likely to also undermine Auer deference, as articulated in Auer v. Robbins, 519 U.S. 452 (1997). The Supreme Court rejected the opportunity to do so in Kisor v. Wilkie, 588 U.S. 558 (2019), and we have no reason to believe that the possibility of overruling Auer would materially affect this case in any way. Auer does not apply here because we are not reviewing an agency’s interpretation of its own regulation. Based on Kisor, however, Rana argues that the State Department’s interpretation must reflect “fair and considered judgment.” Id. at 579. 6 Even if Auer did apply, the State Department’s technical analysis need not “be supported with reasoning and analysis,” as Rana contends, because it is merely a recounting of the Executive Branch’s understanding of the treaty it negotiated. Rana refers us to Hill v. Norton, 275 F.3d 98 (D.C. Cir. 2001), as persuasive authority that we need not defer to the Executive’s interpretation of treaties. In that case, the D.C. Circuit held that the failure of the Secretary of the Interior to include a certain bird species on a list of birds protected by the Migratory Bird Treaty Act was arbitrary and capricious under the APA. Id. at 99. That case, however, is distinct for two reasons. First, Hill involved review of post-ratification implementation of a treaty via regulations, not a technical analysis explaining the intent of the drafters. Id. The Secretary of the Interior’s regulations did not evince a pre- ratification intent. Second, the Secretary’s implementing regulation contradicted the clear, unambiguous terms of the treaty, which is, as stated above, not the case here. See id. at 106. The plain meaning of the Treaty is supported by the Executive’s understanding of its terms at the time of drafting. We defer to that understanding.

6 Rana also argues that we could only defer to the State Department’s interpretation of an ambiguous treaty. We come to our conclusion primarily based on the unambiguous plain text of the Treaty, so this argument makes no difference.

RANA V. JENKINS 17

C. Persuasive case law supports an elements approach. The Fourth Circuit considered a similar treaty provision in Zhenli Ye Gon v. Holt (Ye Gon), 774 F.3d 207 (4th Cir. 2014), and required an elements-based analysis. The Non Bis in Idem provision there stated that the requested country shall not extradite a fugitive who “‘has been prosecuted or has been tried and convicted or acquitted’ in that country, if that prosecution or trial was ‘for the offense for which extradition is requested.’” Id. at 211. The Fourth Circuit interpreted “offense” to mean “the definition of the crime.” Id. at 215. The Fourth Circuit compared the Non Bis in Idem provision to the dual criminality provision of the treaty. Id. The dual criminality provision prevented extradition for crimes unless they were “wilful acts . . . punishable in accordance with the laws of both Contracting Parties.” Id. Because the dual criminality provision used the word “acts” and the Non Bis in Idem provision used the word “offense[s],” the Fourth Circuit held that “offenses” “must be something other than the acts underlying those offenses.” Id. The most natural reading of “offenses,” the Fourth Circuit explained, is the definition of the crime, supporting the double jeopardy approach outlined in Blockburger v. United States, 284 U.S. 299 (1932). Ye Gon, 774 F.3d at 215. Although the dual criminality provision here also uses the word “offense,” it explicitly limits “offense” to mean conduct for that provision only. See Knotek, 925 F.3d at 1131 n.12. Here, the word “acts” is used in the very next paragraph in Article 6, leading to an even stronger inference that “acts” and “offense[s]” have distinct definitions. See Air France, 470 U.S. at 398. Similarly, the Eleventh Circuit in United States v. Duarte-Acero, 208 F.3d 1282 (11th Cir. 2000), read the word “offense” in a Non Bis in Idem provision to refer narrowly to criminal elements, as opposed to “[a] broader formulation[,] [which] would have used the word ‘act,’ ‘action,’ or ‘conduct.’” Id. at 1286; see also United States v. Trabelsi, 845 F.3d 1181, 1193 (D.C. Cir. 2017) (Pillard, J., concurring in part and concurring in the judgment) (endorsing a Blockburger analysis in interpreting whether the same “offense” was prosecuted in each country). Rana also refers us to Sindona v. Grant, 619 F.2d 167 (2d Cir. 1980), in which the Second Circuit interpreted “offense” to mean underlying conduct. That decision, however, is less persuasive for several reasons. First, the “same conduct” test in Sindona was, at least in part, based on a concurrence by Justice Brennan in Ashe v. Swenson, 397 U.S. 436 (1970), which the Supreme Court eroded in United States v. Dixon, 509 U.S. 688, 704 (1993) (striking down the “same conduct” rule for double jeopardy analysis as “wholly inconsistent with earlier Supreme Court precedent and with the clear common-law understanding of double jeopardy”). See Ye Gon, 774 F.3d at 216–17. Second, Sindona relied on the policy of the Department of Justice not to try federal cases where a state has prosecuted “the same act or acts.” 619 F.2d at 178 (citing Petite v. United States, 361 U.S. 529, 530–31 (1960) (per curiam)). The internal DOJ Petite Policy has no bearing on what the word “offense” means in the Treaty, particularly where the policy does not use the word. See Ye Gon, 774 F.3d at 216. Finally, Sindona relied on the argument that a foreign country “could hardly be expected to be aware of Blockburger.” 619 F.2d at 178. While that may have been true in 1980, the Treaty was signed nearly twenty years later in 1997. Even if India were not aware of

RANA V. JENKINS 19

Blockburger, our conclusion that “offense” and “acts” must have different meanings stands. Rana has offered no other non-elements-based definition of “offense” that would distinguish it from Article 6’s use of “acts” in Paragraph 2. Thus, nothing in Sindona persuades us that “offense” does not refer to a crime’s elements. 7 D. Headley’s plea agreement does not compel a different result. Rana contends that the government’s interpretation of Article 6 in Headley’s plea agreement demonstrates that “offense” refers to conduct rather than elements. Headley’s plea agreement states: “Pursuant to Article 6 of the Extradition Treaty . . . defendant shall not be extradited . . .

7 Along similar lines, Rana suggests that the government’s interpretation of “offense” creates absurd results because there may be local or country- specific elements included in a crime, such as an effect on interstate commerce. The plain language of the Treaty, however, suggests that this may be the intended result. In Article 2, Section 3(b), the Treaty states that for purposes of the dual criminality provision, an offense is extraditable “whether or not the offense is one for which United States federal law requires the showing of such matters as interstate transportation, or use of the mails or of other facilities affecting interstate or foreign commerce, such matters being merely for the purpose of establishing jurisdiction in a United States federal court[.]” The inclusion of this provision in Article 2 demonstrates that the drafters were aware of difficulties in comparing United States and Indian law but chose not to include such an exception in Article 6. In any event, neither of the terrorism conspiracies for which Rana was prosecuted under 18 U.S.C. § 2339A required proof of a local element. If, at an appropriate time, we determine it is necessary to carve out a jurisdictional-element exception to Blockburger to apply exclusively in extradition cases, we may do so. Cf. United States v. Hairston, 64 F.3d 491, 496 (9th Cir. 1995). At this time, we take no stance on whether such an exception would be appropriate or necessary. for any offenses for which he has been convicted in accordance with this plea.” The next sentence states:

The defendant and the United States Attorney’s Office accordingly agree that, if defendant pleads guilty to and is convicted of all offenses set out in the Superseding Indictment, including Conspiracy to Bomb Places of Public Use in India (in violation of 18 U.S.C. § 2332f(a)(2)) . . . then the defendant shall not be extradited to the Republic of India . . . for the foregoing offenses, including conduct within the scope of those offenses for which he has been convicted . . . (emphasis added).

Rana further points to statements made by U.S. Attorney Patrick J. Fitzgerald at Headley’s plea colloquy describing the above plea provision:

Your Honor, I think that is a summary of what is a paragraph that is very specific in Paragraph 9 of the agreement . . . . And that says if the conduct is conduct within the scope of those offenses for which he has been convicted in accordance with the plea, then according to the treaty, he would not be extradited.

Taking the plea agreement by itself, the language does not obviously suggest that “offense” means “conduct.” The plea agreement specifies that, for purposes of the agreement, offense “include[s] conduct within the scope of those offenses,” thus suggesting that “offense” is not co-extensive

RANA V. JENKINS 21

with “conduct” in every case. Additionally, the “pursuant to Article 6” language occurs only in the prior sentence, which refers exclusively to offenses. The plea agreement then goes on to list the “offenses” for which Headley had been indicted, all of which refer to specific statutory criminal violations. The U.S. Attorney’s comments are less clear. Although they do seem to suggest that he interpreted the plea agreement to employ a conduct-based test under the Treaty, Rana provides no case that suggests that a singular statement by a U.S. Attorney interpreting a plea agreement should control over both plain language in the Treaty and the earlier-in-time contemporaneous understanding of the State Department. Instead, Rana asks this court to judicially estop the Justice Department from changing positions between Headley’s and Rana’s proceedings. He also asks that, in the alternative, we remand to the district court for production and examination of internal Justice Department communications. A party is judicially estopped from making an argument “when 1) its current position is ‘clearly inconsistent’ with its previous position; 2) ‘the party has succeeded in persuading a court to accept that party’s earlier position’; and 3) the party, if not estopped, ‘would derive an unfair advantage or impose an unfair detriment on the opposing party.’” Perez v. Discover Bank, 74 F.4th 1003, 1008 (9th Cir. 2023). We decline to estop the government here because, even if U.S. Attorney Fitzgerald’s statement was “clearly inconsistent” with the interpretation it now advances, it fails to meet the latter two prongs. First, the government did not persuade the district court to adopt its prior interpretation of the Treaty; it merely asked the district court to approve a plea agreement that contained a provision that may have been broader than the Treaty. The district court did not explicitly accept any interpretation of the Treaty in approving the agreement. Second, Rana does not successfully articulate why the government would derive an unfair advantage from pursuing an interpretation of the Treaty that is supported by its plain language and technical analysis. Thus, we also decline to remand the case to the district court for production and examination of internal Justice Department communications. We conclude that “offense” in Article 6, Paragraph 1 requires us to compare the elements of the crime for which Rana was acquitted in the United States with the elements of the crimes he is charged with in India. Because the parties do not dispute that the crimes charged in India have elements independent from those under which Rana was prosecuted in the United States, the Treaty permits Rana’s extradition.

II. Competent evidence supports probable cause.

Article 9.3(c) of the Treaty provides that a request for extradition must be supported by “such information as would justify the committal for trial of the person if the offense had been committed in the Requested State.” The parties agree that this standard requires India to provide information “that would be sufficient to establish probable cause” that Rana committed the alleged crime. Emami v. U.S. Dist. Court for N. Dist. of Cal., 834 F.2d 1444, 1447 (9th Cir. 1987). We must affirm the probable cause finding so long as “there was any evidence warranting the finding that there was reasonable ground to believe the accused guilty.” Manrique, 65 F.4th at 1044.

RANA V. JENKINS 23

Rana attacks the credibility of Headley, who was the government’s main witness. Rana argues that the extradition court should have discounted Headley’s testimony because he (a) is a serial cooperator who returned to criminal activity, (b) “received training in manipulation and deception by the ISI, Pakistan’s intelligence service,” (c) initially denied Rana’s knowledge of Headley’s terrorist activities and then only disclosed Rana’s knowledge after Rana’s arrest, (d) used Rana for illegal activity without Rana’s knowledge, (e) deceived his multiple wives, (f) and told other unrelated falsehoods to his associates and a judge. Questions concerning Headley’s credibility, however, are not properly before us. “An accused in an extradition hearing has no right . . . to pose questions of credibility as in an ordinary trial . . . .” Santos v. Thomas, 830 F.3d 987, 992 (9th Cir. 2016) (en banc) (quoting Eain v. Wilkes, 641 F.2d 504, 511 (7th Cir. 1981)). The only evidence an accused can introduce “is evidence that ‘explains away or completely obliterates probable cause.’” Id. (quoting Mainero v. Gregg, 164 F.3d 1199, 1207 n.7 (9th Cir. 1999), superseded by statute on other grounds as recognized in Manrique, 65 F.4th at 1044). Attacks on Headley’s credibility, while perhaps compelling to a jury, do not rise to the level of “complete[] obliterat[ion]” required to find a lack of probable cause. Id. In Barapind v. Enomoto, for example, our court sitting en banc rejected the accused’s proffer of a “significant” government witness’s complete recantation of his identification of the accused. 400 F.3d 744, 749–50 (9th Cir. 2005) (per curiam) (en banc). None of the attacks Rana proffers regarding Headley’s credibility comes anywhere close to a direct recantation of his testimony, and all of them would require evaluations of credibility best left for trial. See Santos, 830 F.3d at 992. The government correctly notes that the cases to which Rana cites largely recite probable cause standards in the criminal context outside of extradition proceedings, see Illinois v. Gates, 462 U.S. 213 (1983); United States v. Jensen, 425 F.3d 698 (9th Cir. 2005); United States v. Nielsen, 371 F.3d 574 (9th Cir. 2004), and it is unclear why those cases apply given the clear, recent precedent proscribing the consideration of impeachment evidence in extradition proceedings. Rana points to Choe v. Torres, 525 F.3d 733 (9th Cir. 2008), to support his argument that this court can consider a credibility challenge to undermine the extradition court’s probable cause finding, but that case did just the opposite. There, we rejected the petitioner’s credibility challenge: “Ho’s lack of credibility is merely a weakness in Korea’s case; it does not ‘completely obliterate[]’ the evidence of probable cause.” Id. at 740. 8 One district court case cited by Rana, In re Extradition of Ameen, 2021 WL 1564520, at *13 (E.D. Cal. Apr. 21, 2021), relies on Quinn v. Robinson, 783 F.2d 776, 815 (9th Cir. 1986), for the proposition that the extradition magistrate judge may consider credibility evidence. Quinn makes clear, however, that our court may not consider credibility attacks on habeas review, even if an extradition magistrate judge may consider them. See id. We review Rana’s habeas petition here, so we may not consider Headley’s credibility. The other out-of-circuit district court case Rana cites, In re Mazur, 2007 WL 2122401 (N.D. Ill. July 20, 2007), is also clearly distinguishable. There, the court evaluated multiple statements surrounding the same event from an individual

8 We do not consider the likelihood that Rana would die in jail or be unable to obtain legal representation in India. We must assume that Rana will face a fair trial. Glucksman v. Henkel, 221 U.S. 508, 512 (1911).

RANA V. JENKINS 25

witness. Id. at *21–22. The court rejected the witness’s statements because the story changed so substantially between statements that it was impossible to believe that the story was true and because the witness “admit[ted] that he lied under oath” and “that he essentially made up the connection between” the petitioner and the criminal activity. Id. Rana has not presented similar evidence here. 9 Thus, no case Rana cites to supports the proposition that this court may consider credibility attacks in determining whether the extradition judge properly found probable cause. Rana’s arguments that the Pakistan Intelligence Service paid for the Mumbai office, and that Rana would not have checked Headley’s false visa application, rely exclusively on an attack on Headley’s credibility, so we reject them. The other evidence that Rana offers (apart from his attack on Headley’s credibility) does not help him. His arguments that (1) the Mumbai office did business, (2) the Mumbai office closed around the time of the attack for legitimate reasons, and (3) the warning about the impending attacks from Headley’s co-conspirator in Dubai disproves Rana’s earlier awareness of the attacks, do not upset the finding of probable cause. First, Rana presents evidence disproving that the “Mumbai office did no business,” but that ignores clear testimony from the business’s customers, stating that they never received the visas for which they paid, and its

9 Rana argues that Headley’s initial refusal to implicate Rana is similar to Mazur, but the extradition court considered Headley’s explanation that he initially lied in an attempt to shield his childhood friend but ultimately offered up the evidence once he learned Rana was arrested. It is not in the province of our court, reviewing a habeas petition, to review that credibility determination. See Quinn, 783 F.2d at 815. secretary, stating that “there [was] no business.” Second, Rana claims that the Mumbai office closing around the time of the attacks does not create an inference that the Mumbai office was a sham. Instead, Rana points to evidence that he was looking for ways to maintain his business presence in Mumbai. Even so, the evidence discussed above is sufficient to demonstrate probable cause that the Mumbai office was a sham. Finally, Rana argues that the fact that Headley sent a third party to warn Rana of the attacks disproves Rana’s involvement with the plot. This inference does not “obliterate” probable cause; an equally compelling inference could be drawn that Headley kept Rana informed of the plans. Competent evidence supports the extradition court’s finding of probable cause.

CONCLUSION

The judgment of the district court is AFFIRMED.