concurring in part and dissenting in part, with whom Judge W. FLETCHER joins:
Hedelito Trinidad y Garcia (“Trinidad”) claims that if extradited to the Philippines, he is more likely than not to be tortured, and that the Secretary of State’s decision to extradite him is therefore unlawful under the Convention Against Torture and the federal statute implementing it, the Foreign Affairs Reform and Restructuring Act of 1998(the FARR Act), Pub.L. No. 105-277, 112 Stat. 2681 (codified at 8 U.S.C. § 1231 note). The per curiam majority opinion holds that: (1) we have jurisdiction to hear Trinidad’s challenge to his extradition; (2) as a matter of due process, the Secretary of State is required to consider Trinidad’s claim that he will be tortured if returned to the Philippines and to refrain from extraditing him if she finds it “more likely than not” that he will indeed be tortured; and (3) without a declaration from the Secretary (or her delegate) that the Secretary has fulfilled her obligation, there is insufficient evidence in the record to determine whether she has done so. I agree. I therefore concur in Parts 1-5 of the majority opinion.
I cannot, however, agree with the majority’s ultimate holding that once the Secretary (or her delegate) meets the procedural due process requirement by submitting a barebones declaration, courts under no circumstances have authority to conduct any substantive review of the Secretary’s compliance with federal law.
*985There is no reason for the majority even to reach this question. Once the majority determines that there has been a procedural due process violation and that therefore “we lack sufficient basis in the record to review the district court’s order granting Trinidad y Garcia’s release,” Per curiam at 957, we should simply remand for the submission of an appropriate declaration. If there is a subsequent appeal, we could then determine whether further substantive review is available and, if so, whether the record is adequate for that purpose.
The majority nevertheless jumps the gun and dismisses Trinidad’s substantive claims, holding, with little explanation, that they are foreclosed by the Supreme Court’s decision in Munaf v. Geren, 553 U.S. 674, 128 S.Ct. 2207, 171 L.Ed.2d 1 (2008), the doctrine of separation of powers, and the “rule of non-inquiry.” Per curiam at 957. Judge Tallman elaborates on these points at length and adds another — the contention that Trinidad has no statute-based claim at all. Tallman dissent at 972-76. I cannot go along with either the majority’s curt conclusion or Judge Tallman’s more discursive analysis. I therefore concur in the majority’s result — a remand to the district court for further development of the record — but not in its declaration that under no circumstances can a district court go further than to require a pro forma declaration from the Secretary of State in a case in which it is alleged that extradition will likely result in torture.
I begin by outlining the basic building blocks of Trinidad’s substantive, statute-based claim.
First, we may grant a writ of habeas corpus where a prisoner is “in custody in violation of the Constitution or laws or treaties of the United States.” 28 U.S.C. § 2241(c)(3).1
Second, Article 3 of the Convention Against Torture (CAT), which entered into force for the United States in 1994, states;
No State Party shall expel, return (“refouter ”) or extradite a person to another state where there are substantial grounds for believing that he would be in danger of being subjected to torture.
United Nations Convention Against Torture and Other Forms of Cruel, Inhuman or Degrading Treatment or Punishment, adopted by unanimous agreement of the U.N. General Assembly, G.A. Res. 39/46, 39 U.N. GAOR Supp. No. 51 at 197, U.N. Doc. A/RES/39/708 (1984), entered into force as to the United States Nov. 20, 1994, signed Apr. 18, 1988. The Senate ratified CAT with the understanding that “the phrase, ‘where there are substantial grounds for believing that he would be in danger of being subjected to torture,’” would be understood to mean “‘if it is more likely than not that he would be tortured.’ ” U.S. Senate Resolution of Advice and Consent to Ratification of the Convention Against Torture and Other Cruel, Inhuman, or Degrading Treatment or Punishment, 136 Cong. Rec. 36, 198 (Oct. 27,1990).
The language of Article 3 is mandatory. Whereas some CAT provisions limit signatories’ obligation to enforce a policy, see, *986e.g., Article 13 (“Steps shall be taken to ensure that ...” (emphasis added)); Article 14 (requiring that signatories provide torture victims “the means for as full rehabilitation as possible ” (emphasis added)), Article 3 has no such equivocation or limitation. Signatories are not, for example, prohibited from expelling individuals likely to face torture “where feasible” or “to the extent possible.” Cf. INS v. Cardozar-Fonseca, 480 U.S. 421, 107 S.Ct. 1207, 94 L.Ed.2d 434 (1987) (distinguishing between a treaty provision that creates an entitlement to be protected from expulsion and one that is discretionary). Instead, the CAT Article 3 prohibition is general and unlimited: Without exception, a signatory country may not extradite a person likely to face torture.
The final building block of Trinidad’s statute-based claim is the FARR Act, which, echoing the language of CAT, provides that:
It shall be the policy of the United States not to expel, extradite, or otherwise effect the involuntary return of any person to a country in which there are substantial grounds for believing the person would be in danger of being subjected to torture....
8 U.S.C. § 1231 note. The FARR Act then directs “the heads of the appropriate agencies” to “prescribe regulations to implement the obligations of the United States.” Id. As the government recognizes in its brief, the FARR Act thereby “prohibits the extradition of a person who more likely than not will be tortured, and ... creates a duty on the part of the Secretary of State to implement that prohibition.” 2
Contrary to Judge Kozinski’s assertion, Trinidad’s claim is not that he is entitled to habeas because of the treatment he is likely to face in the Philippines. Rather, his claim is a claim that because the FARR Act prohibits extradition if, on the information available to the Secretary, he more likely than not will be tortured, the Secretary’s decision to extradite him would be illegal under positive, Congressionally enacted federal law. In other words, the focus of Trinidad’s habeas petition is on the legality of the Secretary’s decision, not on whether or not Trinidad will actually be tortured if extradited. This claim is one at the “historical core” of habeas review. INS v. St. Cyr, 533 U.S. 289, 301, 121 S.Ct. 2271, 150 L.Ed.2d 347 (2001). Indeed, it is “as a means of reviewing the legality of Executive detention ... that [the] protections [of the writ of habeas corpus] have been strongest.” Id.; see generally Gerald L. Neuman, The Habeas Corpus Suspension Clause After Boumediene v. Bush, 110 Colum. L. Rev. 537, 541 (2010).
As I explain below, neither the Supreme Court’s decision in Munaf nor the rule of non-inquiry entirely forecloses our ability to review the lawfulness of an extradition decision by the Executive. I would hold, therefore, that we have the authority— and, indeed, the obligation — to review the Secretary of State’s determination and to decide — under a standard highly deferential to the Secretary and procedures carefully tailored to ensure the protection of the Secretary’s diplomatic concerns— whether it is more likely than not that petitioners such as Trinidad will be tortured if extradited. For that purpose, it may be that in many circumstances a dec*987laration such as the one the majority requires will suffice. But, as I shall explain below, not invariably.
I. The FARR Act
Before doing so, however, I address a separate proposition put forth by Judge Tallman but not addressed by the majority. Judge Tallman maintains that despite the Government’s emphatic assertion to the contrary, the FARR Act does not actually restrict the Executive’s discretion to extradite, even when it is more likely than not that an individual will be tortured. Instead, Judge Tallman insists, the FARR Act is merely “precatory”; it serves no other purpose than to “ ‘nudge’ ” the government in Congress’s “ ‘preferred direction ].’ ” Tallman dissent at 977 (quoting Pennhurst State Sch. & Hosp. v. Halderman, 451 U.S. 1, 19, 101 S.Ct. 1531, 67 L.Ed.2d 694 (1981)). This understanding of the Act could not be more wrong.
Judge Tallman’s argument proceeds from his reading of section (a) of the FARR Act. According to Judge Tallman, that section, which states that “[i]t shall be the policy of the United States not to ... extradite ... any person to a country in which there are substantial grounds for believing the person would be in danger of being subjected to torture,” 8 U.S.C. § 1231 note, only announces a general policy of the United States, imposing no obligation on the Executive to comply in any specific instance. For this proposition, he relies on Pennhurst.
Pennhurst concerned whether the statement of congressional findings included in the federal Developmentally Disabled Assistance and Bill of Rights Act imposed upon the states an obligation to fund particular kinds of mental healthcare entitlements. For several reasons, the case is entirely inapposite here.
For one thing, the statute at issue in Pennhurst was passed either pursuant to Congress’s power under § 5 of the Fourteenth Amendment or pursuant to its spending power.3 The Supreme Court has held that statutes passed pursuant to either of these powers and intended by Congress to impose obligations on the states must clearly state this intention, particularly where the obligation is the creation of an affirmative entitlement. See id. at 16-18, 101 S.Ct. 1531. The FARR Act, however, binds only the federal government; it does not purport to impose any obligations upon the states. As a result, the federalism concerns animating Pennhurst simply do not apply here, and no clear statement rule of the kind applied in Pennhurst applies to this case.
In addition, the Supreme Court found the language at issue in Pennhurst ambiguous as to whether it imposed an obligation upon the states enforceable by individuals. The Court therefore turned to the remainder of the statute to determine whether, in context, the import of the ambiguous provision became clear. See id. at 19, 101 S.Ct. 1531. The disputed congressional “findings” in Pennhurst were embedded in a statute, other sections of which clearly and explicitly imposed obligations on the states. See id. These specific obligations would have been redundant were the more general “findings” in the statute considered binding commands. See id. at 19, 25-27, 101 S.Ct. 1531. The FARR Act contains no analogous specific provisions.
In fact, consistent with the Government’s position, the text and structure of the FARR Act confirm that it does impose *988a binding obligation on the Secretary of State not to extradite individuals likely to face torture. Subsection (a) of the FARR Act incorporates the language of CAT itself, enacting as U.S. domestic policy the international obligation the United States undertook in ratifying CAT. See 8 U.S.C. § 1281 note. The remainder of the Act then directs the Executive “to implement the obligations of the United States under” CAT and specifies how such implementation ought to occur. Id. Whereas the statute at issue in Pennhurst combined an aspirational vision for the ideal treatment of people with disabilities with more specific mandates, there is nothing aspirational about the FARR Act. It states a policy and directs agencies to implement that policy. If this policy is merely precatory, then all of the FARR Act would also be so. I cannot agree that Congress passed a statute with no intent to affect anyone’s rights or obligations.
Judge Tallman, however, reads the FARR Act’s incorporation of CAT differently, maintaining that the Act’s direction in subsection (b) that “the heads of appropriate agencies shall prescribe regulations to implement” the United States’ “obligations” under CAT, id., “conflicts with Trinidad’s assertion that the FARR Act itself implements the Convention and binds Executive authority.” Tallman dissent at 975. Trinidad’s assertion, however, is that the FARR Act implements CAT, and so makes the United States’ obligations under CAT binding not only as a matter of international law — as they became when the United States signed CAT — but as a matter of U.S. law. The FARR Act’s mandate to agencies that they “implement” the United States’ obligations under CAT is a direction to put into practice the mandatory Article 3 obligations undertaken by signing CAT and incorporated into U.S. law by the FARR Act. That mandate would be absurd if, as Judge Tallman insists, no such obligations exist under U.S. law at all.
Further, assuming subsection (a) does no more than express a general policy, subsection (b) of the FARR Act is unquestionably obligatory. Thus, even disregarding entirely subsection (a), subsection (b) compels the conclusion that the FARR Act imposes upon the Executive an obligation to abide by CAT.
As a fallback to his insistence that the FARR Act is simply precatory and does not bind the executive at all, Judge Tall-man more modestly proposes that subsection (d) of the Act demonstrates that “Congress did not intend to impose an obligation on the Executive outside the removal context.” Tallman dissent at 6435-36. In other words, Judge Tallman suggests a cleavage in the substantive duty created by the Act between the Executive’s obligation in the immigration removal context and that applicable in all other circumstances, including extradition. This more narrow contention fares no better than Judge Tallman’s broader, Pennhurstgrounded one.
The problem with this alternative suggestion is that there is no indication whatsoever in the statute that the substantive obligations it imposes vary by context. Subsection (d), on which Judge Tallman relies for his contrary proposition, describes only courts’ authority to review FARR Act claims, not the substantive reach of the underlying governmental obligation. That the FARR Act specifically allows for jurisdiction to review claims in the removal context but leaves review in any other context dependent on pre-existing jurisdiction (as Judge Tallman recognizes in his jurisdictional analysis) does not alter the substance of the obligation the Act creates. That obligation, the imposition of a uniform policy prohibiting “the involuntary return of any person to a coun*989try” where the person likely to face torture, 8 U.S.C. § 1281 note, is affirmatively stated and generally applicable.
If anything, the inclusion of the provision addressing courts’ jurisdiction to review FARR Act claims further supports the view that the Act creates obligations out of which claims could arise. A provision allowing for the review of FARR Act claims as part of the review of final orders of removal would be meaningless if no such claims could ever arise because the FARR Act created no governmental duty ■with regard to expelling individuals facing torture.
Judge Tallman’s last stab at finding a basis for declaring that, despite the FARR Act, the Government still has discretion to extradite a detainee facing torture is a State Department regulation providing that the Secretary’s extradition decisions “are matters of executive discretion not subject to judicial review.”4 22 C.F.R. § 95.4. But the parties, the Supreme Court, and the courts of appeals have all taken the view that the FARR Act implements CAT by incorporating the obligations undertaken in the treaty into domestic law, thereby eliminating any discretion the Secretary of State might otherwise have had to extradite a person likely to face torture. The State Department, in particular, not only, as noted, agrees in its brief with Trinidad’s contention that “the FARR Act creates a duty on the part of the Secretary of State to implement” the Act’s “prohibition” against extraditing “a person who more likely than not will be tortured,” but further assures us that it “is not arguing that the Secretary of State has discretion to surrender a fugitive who more likely than not will be tortured, even if foreign policy interests at the time would be served by an extradition.” The Secretary’s own interpretation of the regulation upon which Judge Tallman relies is clearly that, whatever discretion the State Department has over extradition decisions, its discretion does not extend to the ability to extradite an individual likely to face torture. This interpretation is controlling. “[W]e defer to an agency’s interpretation of its own regulation, advanced in a legal brief, unless that interpretation is ‘plainly erroneous or inconsistent with the regulation.’ ” Chase Bank USA, N.A. v. McCoy, — U.S. -, 131 S.Ct. 871, 880, 178 L.Ed.2d 716 (2011) (quoting Auer v. Robbins, 519 U.S. 452, 461, 117 S.Ct. 905, 137 L.Ed.2d 79 (1997)). Given the Secretary’s contrary view, Judge Tallman’s reading of the State Department regulations as providing discretion with regard to FARR Act obligations cannot stand.
Consistent with the Government’s understanding, the Supreme Court, in Medellin v. Texas, cited the FARR Act as exemplifying a statute by which a treaty (CAT) had been given “wholesale effect ... through implementing legislation.” Medellin v. Texas, 552 U.S. 491, 520, 128 S.Ct. 1346, 170 L.Ed.2d 190 (2008). Were we to hold that the FARR Act did not, in fact, implement as domestic law the obligations undertaken in CAT, but only “nudged” the Executive toward refraining from sending persons abroad to face torture, we would be contradicting the view expressed — albeit in dicta — by the Supreme Court.
We would also be overruling several of our own circuit’s cases. See, e.g., Edu v. Holder, 624 F.3d 1137, 1144 (9th Cir.2010) (“Congress then implemented CAT in the *990Foreign Affairs Reform and Restructuring Act of 1998.”); Huang v. Ashcroft, 390 F.3d 1118, 1121 (9th Cir.2004) (“Congress passed the Foreign Affairs Reform and Restructuring Act (the FARR Act) in 1998 to implement Article 3 of CAT.”); Zheng v. Ashcroft, 332 F.3d 1186, 1193 (9th Cir.2003) (“In 1998, Congress passed the Foreign Affairs Reform and Restructuring Act of 1998, implementing Article 3 of the Convention Against Torture.”). Furthermore, we would be contradicting the law of our sister circuits, none of which has doubted that the FARR Act implements CAT. See, e.g., Omar v. McHugh, 646 F.3d 13, 18 n. 2 (D.C.Cir.2011) (“[I]t is undisputed that the FARR Act implements the Convention Against Torture.”); Pierre v. Att’y Gen., 528 F.3d 180, 185-86 (3d Cir.2008) (“[I]n 1998, Congress passed legislation to implement the United States’ obligations under the CAT: the Foreign Affairs Reform and Restructuring Act (‘the FARR Act’).”); Pierre v. Gonzales, 502 F.3d 109, 114 (2d Cir.2007) (“To implement the CAT, Congress amended the immigration laws with the Foreign Affairs Reform and Restructuring Act of 1998 (‘the FARR Act’).”); Cadet v. Bulger, 377 F.3d 1173, 1180 (11th Cir.2004) (“In order to implement Article 3 of CAT, Congress passed the Foreign Affairs Reform and Restructuring Act of 1988 (‘the FARR Act’).”).
There is simply no doubt that as a sub-' stantive matter, the FARR Act imposes a binding obligation on the Secretary of State not to extradite a person likely to face torture. The majority agrees with this proposition. Per curiam at 956-57.
II. Munaf v. Geren
The majority does maintain that once the Secretary provides a declaration stating that she complied with her CAT and FARR Act obligations, Munaf (and the “rule of non-inquiry,” which I address in due course) preclude judicial inquiry in any and all circumstances — even if, for example, there was irrefutable evidence that torture was indeed more likely than not to occur were the detainee to be extradited. Per curiam at 957. I cannot agree.
Munaf does not foreclose, or even very much affect, our authority to review Trinidad’s FARR Act claim. To begin, Munaf emphatically declined to decide the question at issue here — whether the FARR Act provides a basis for habeas review of the Secretary of State’s extradition decisions. That Munaf reserved rather than decided the question before us could not be more clear. The Court stated:
Petitioners briefly argue that their claims of potential torture may not be readily dismissed ... because the FARR Act prohibits transfer when torture may result. Neither petitioner asserted a FARR Act claim in his petition for habeas, and the Act was not raised in any of the certiorari filings before this Court. Even in their merits brief in this Court, the habeas petitioners hardly discuss the issue. The Government treats the issue in kind. Under such circumstances we will not consider the question.
Munaf, 553 U.S. at 703, 128 S.Ct. 2207 (internal citations omitted and emphasis added).
This reservation in Munaf is of more than technical import. It indicates that there at least could be some difference of controlling significance between a claim based on an affirmative Congressional enactment, like the FARR Act, placing obligations on the Executive Branch and a constitutionally based norm, such as the one invoked in Munaf. Had it been clear that there is no such possible difference, as the majority opinion in this case tacitly assumes, then there would have been no reason to reserve the question.
*991Moreover, not only is there no applicable holding in Munaf; there is no applicable reasoning or implicit “message” either. Instead, the Court’s reasoning in Munaf is tightly bound to the factual and legal circumstances in which the case arose, see id. at 700, 128 S.Ct. 2207 (characterizing its analysis as applying “in the present context”), circumstances that differ completely from those here.
Most notably, Munaf was not an extradition case. The Munaf petitioners were detained in Iraq, at the request of the Iraqi government, by an international military coalition commanded by the United States. The charges against them were that they violated Iraqi criminal law. They sought not the traditional habeas remedy of release from executive detention but, rather, affirmative protection from the reach of the Iraqi government. As the Munaf Court explained:
[T]he nature of the relief sought by the habeas petitioners suggests that habeas is not appropriate in these cases. Habeas is at its core a remedy for unlawful executive detention. The typical remedy for such detention is, of course, release. But here the last thing petitioners want is simple release; that would expose them to apprehension by Iraqi authorities for criminal prosecution— precisely what petitioners went to federal court to avoid. At the end of the day, what petitioners are really after is a court order requiring the United States to shelter them from the sovereign government seeking to have them answer for alleged crimes committed within that sovereign’s borders.
Id. at 693-94, 128 S.Ct. 2207; see also id. at 704, 128 S.Ct. 2207 (“Omar and Munaf voluntarily traveled to Iraq and are being held there. They are therefore subject to the territorial jurisdiction of that sovereign, not of the United States.... It would be more than odd if the [U.S.] Government had no authority to transfer them to the very sovereign on whose behalf, and within whose territory, they are being detained.”). The relief the Munaf petitioners sought was thus farther from the historical remedy available under habeas than the relief Trinidad seeks — simple release from custody — and more deeply implicated issues of national sovereignty and international comity.
I make this distinction not to suggest that there are not real foreign affairs and international comity concerns in ordinary extradition cases such as this one. See Part IV infra. But these concerns simply do not rise to the level of those at issue in Munaf. While significant, the foreign affairs and comity concerns in the present circumstances are manageable through appropriately deferential habeas procedures and limitations on the scope of judicial review, as I suggest below.
Furthermore, the Munaf petitioners’ claims raised military and national security concerns that Trinidad’s claims do not. At least one of the Munaf petitioners was charged with terrorism-related crimes. And the Court repeatedly emphasized that the case took place “in the context of ongoing military operations.” Id. at 689, 128 S.Ct. 2207.
Moreover, and critically for present purposes, Munaf affirmatively left open not only the FARR Act issue but also the question of whether the result could be different in “a more extreme case in which the Executive has determined that a detainee is likely to be tortured but decides to transfer him anyway.” Id. at 702, 128 S.Ct. 2207. Justice Souter, joined by Justices Ginsburg and Breyer, saw a wider opening; they “would extend the caveat to a case in which the probability of torture is well documented, even if the Executive fails to acknowledge it.” Id. at 706, 128 S.Ct. 2207 (Souter, J., concurring). The *992majority here, however, closes the door Munaf opened, ruling that once a detainee in Trinidad’s position is afforded the procedural assurance that the Secretary has considered her CAT obligations, there is no substantive review whatsoever available, constitutional or statutory, no matter what the underlying circumstances.
I conclude that given Munaf & refusal to answer the question presented in this case, as well as the substantial differences between the two cases, Munaf is of little use here.
III. The Rule of Non-Inquiry
The majority’s, and Judge Tallman’s, more basic ground for shutting the door on any judicial consideration of Trinidad’s substantive claims is the rule of non-inquiry. Consideration of the background and role of that principle in extradition cases demonstrates that it does not apply — at least without substantial adjustment— where, as here, there is a specific, mandatory directive to the Executive Branch with regard to the treatment of extradition requests.
There is no constitutional or statutory command establishing a rule of non-inquiry — that is, a rule precluding any substanfive judicial inquiry into the likely fate of extradited criminal defendants.5 Rather, the traditional principle that has been dubbed the “rule of non-inquiry” developed as a judge-made doctrine, under which “[a]n extraditing court will generally not inquire into the procedures or treatment which await a surrendered fugitive in the requesting country.” Arnbjornsdottir-Mendler v. United States, 721 F.2d 679, 683 (9th Cir.1983).
The Supreme Court has never used the term “rule of non-inquiry,” let alone explicated its scope or proper application. Instead, the doctrine developed “by implication,” as lower courts interpreted and expounded upon Supreme Court extradition precedents. See Semmelman, supra at 1211-12; Mironescu v. Costner, 480 F.3d 664 (4th Cir.2007); John T. Parry, International Extradition, the Rule of Non-Inquiry, and the Problem of Sovereignty, 90 B.U.L. Rev. 1973, 1978-96 (2010).
Since the late nineteenth century, extradition has been a bifurcated process, with the initial determination of extraditability assigned by statute to a magistrate,6 see 18 U.S.C. § 3184, and the final decision to the *993Secretary of State. In elaborating the rule of non-inquiry, courts have relied on two strands of late nineteenth and early twentieth century extradition caselaw.
The first strand of rule of non-inquiry jurisprudence arises out of a series of cases in which the Supreme Court articulated the extradition issues subject to review by a habeas court when examining a magistrate’s decision certifying extraditability. See, e.g., Fernandez v. Phillips, 268 U.S. 311, 45 S.Ct. 541, 69 L.Ed. 970 (1925); Oteiza v. Jacobus, 136 U.S. 330, 10 S.Ct. 1031, 34 L.Ed. 464 (1890); Benson v. McMahon, 127 U.S. 457, 8 S.Ct. 1240, 32 L.Ed. 234 (1888). At issue in these early cases was the procedure and evidence of guilt required before a magistrate could issue a certificate of extraditability.
The Court’s initial cases in this line established that an extradition proceeding is not analogous to a criminal trial, “by which the prisoner could be convicted or acquitted of the crime charged against him,” but is more like a preliminary hearing “for the purpose of determining whether a case is made out which will justify the holding of the accused” for trial. Oteiza, 136 U.S. at 334-35, 10 S.Ct. 1031; see also Benson, 127 U.S. at 462, 8 S.Ct. 1240. The scope of review by a habeas court is correspondingly narrow: On habeas, courts need and ought not issue “a writ of error,” examining all possible procedural defects of an extradition proceeding, Oteiza, 136 U.S. at 334, 10 S.Ct. 1031, for “[f]orm is not to be insisted upon beyond the requirements of safety and justice,” Fernandez, 268 U.S. at 312, 45 S.Ct. 541 (internal citation omitted). Instead, it is sufficient that there be “[c]ompetent evidence to establish reasonable grounds” for extradition; the evidence (and the procedure used to evaluate it) need not be “competent to convict.” Id. Habeas review of a magistrate’s extradition decision, then, “is available only to inquire whether the magistrate had jurisdiction, . whether the offence charged is within the treaty and, by a somewhat liberal extension, whether there was any evidence warranting the finding that there was reasonable ground to believe the accused guilty.” Id.
From this language, Judge Tallman concludes that judicial review in all extradition cases is limited to such a narrowly circumscribed examination of a magistrate’s finding of extraditability and of the magistrate’s jurisdiction to enter such a finding. This position rests on a misunderstanding of the Court’s earliest extradition cases. Those cases, as I have explained, dealt solely with challenges to the extradition proceedings held before a magistrate and were designed only to ensure that there was some basis for the extradition request. There was no claim in these cases that, for example, the Secretary of State’s decision to extradite was contrary to law or, in particular, that the petitioner would face torture if extradited. Given their narrow purview, this line of magistrate review cases does not broadly limit the kinds of claims that may be brought to contest extradition or delimit the scope of judicial review with respect to all such claims. Rather, Fernandez and similar cases established the scope of review for one particular kind of claim — a claim that the magistrate’s decision to certify extraditability was improper. In other words, as the Seventh Circuit has recognized, “these references [to limited review of extradition decisions] ... have occurred in cases that have involved challenges to the findings of the magistrate in the magistrate’s certification hearing and have not involved constitutional challenges to the conduct of the executive branch in deciding to extradite the accused.” In re Burt, 737 F.2d 1477, 1483 (7th Cir.1984).
Exemplifying the second strand of Supreme Court cases from which the federal courts have derived the rule of non-inquiry *994is Neely v. Henkel, 180 U.S. 109, 21 S.Ct. 302, 45 L.Ed. 448 (1901), upon which Judge Tallman heavily relies. In Neely, Charles Neely, accused of embezzling public funds while serving as Finance Agent of the Department of Posts in Cuba, challenged the constitutionality of the statute governing extradition. The statute, Neely argued, “d[id] not secure to the accused, when surrendered to a foreign country for trial in its tribunals, all of the rights, privileges and immunities that are guaranteed by the Constitution to persons charged with the commission in this country of crime against the United States.” Id. at 122, 21 S.Ct. 302. As a consequence, Neely maintained, the federal courts had the authority and responsibility to declare the statute invalid and order his release. The Court rejected this argument, explaining that the constitutional provisions cited by Neely, those
relating to the writ of habeas corpus, bills of attainder, ex post facto laws, trial by jury for crimes, and generally to the fundamental guarantees of life, liberty and property.... have no relation to crimes committed without the jurisdiction of the United States against the laws of a foreign country.
.... When an American citizen commits a crime in a foreign country he cannot complain if required to submit to such modes of trial and to such punishment as the laws of that country may prescribe for its own people, unless a different mode be provided for by treaty stipulations between that country and the United States.
Id. at 123, 21 S.Ct. 302 (emphasis added).
Essentially, then, Neely expresses the noncontroversial proposition that the United States Constitution does not bind other nations. Trinidad’s claim that he will face torture if extradited is superficially similar to Neely’s claim that he would face an abrogation of his constitutional rights, privileges, and immunities upon extradition to Cuba; both are claims about the treatment an extraditee is likely to face in the requesting country. Importantly, however, Trinidad does not claim that he has a right under the U.S. Constitution not to be tortured in the Philippines by Philippine officials. Rather, Trinidad’s claim is based on an affirmative Congressional enactment that enforces a treaty obligation— which Neely recognizes may be subject to domestic enforcement — and that, as I have shown, binds U.S. government officials and prohibits them from extraditing persons likely to be tortured. So, in this case, the issue is not whether foreign officials may be bound by U.S. norms, or about whether the judiciary, rather than the Executive Branch, can enforce constitutional norms with regard to extradition requests. Instead, the question here concerns the role of the judiciary in enforcing the statutory obligations affirmatively placed upon U.S. officials by Congress. In other words, Judge Kozinski’s argument notwithstanding, the claim is not — or, at least, not solely — about Trinidad’s rights once extradited, but rather about the legitimate scope of executive authority — and, in particular, compliance with Congressional limits on that authority designed to ensure compliance with treaty obligations.
This difference is not merely semantic. Whatever authority we may have to review claims that an individual ought not be extradited because of conditions in the receiving country,7 we certainly have the au*995thority and the responsibility to review the legality of executive detention. While the judiciary may not evaluate the constitutionality of the conduct of foreign governments, it is indubitably the role of courts to ensure that American officials obey the law. Indeed, courts have repeatedly declined to apply the rule of non-inquiry to claims that the Executive has acted unlawfully. See, e.g., Mironescu v. Costner, 480 F.3d 664, 670-71 (4th Cir.2007); In re Burt, 737 F.2d at 1483; Plaster v. United States, 720 F.2d 340, 348(4th Cir.1983); see also Parry, supra, at 1998-99.
The Executive’s authority to extradite is neither inherent nor unlimited. Rather, “[i]n the extradition context, when a ‘fugitive criminal’ is found within the United States, ‘there is no authority vested in any department of the government to seize [him] and surrender him to a foreign power’ in the absence of a pertinent constitutional or legislative provision.” Munaf, 553 U.S. at 704, 128 S.Ct. 2207 (quoting Valentine v. United States, 299 U.S. 5, 9, 57 S.Ct. 100, 81 L.Ed. 5 (1936)). As “[tjhere is no executive discretion to surrender [a person] to a foreign government, unless that discretion is granted by law,” Valentine, 299 U.S. at 9, 57 S.Ct. 100, that discretion is circumscribed by the terms of such delegation.
Habeas review in this statutory context, then, does not violate separation of powers principles. On the contrary, it prevents the inappropriate concentration of power within a single branch, where that branch has been assigned mandatory obligations. Judicial review of compliance with Congress’s directives concerning extradition preserves “the delicate balance of governance,” ensuring that the executive’s discretion to extradite is exercised within the parameters of the law established by Congress. See Hamdi v. Rumsfeld, 542 U.S. 507, 536, 124 S.Ct. 2633, 159 L.Ed.2d 578 (2004). The preservation of this separation of powers serves to secure individual liberty, preventing the extradition of those likely to face torture. “Security subsists, too, in fidelity to freedom’s first principles. Chief among these are freedom from arbitrary and unlawful restraint and the personal liberty that is secured by adherence to the separation of powers. It is from these principles that the judicial authority to consider petitions for habeas corpus relief derives.” Boumediene v. Bush, 553 U.S. 723, 797, 128 S.Ct. 2229, 171 L.Ed.2d 41 (2008).
Judge Tallman argues that, whatever the limits on the executive’s authority to extradite, we do not have the power to review compliance with those limitations, unless specifically authorized by Congress. Indeed, on Judge Tallman’s view, we may not review any aspect of extradition absent specific congressional authorization. The Supreme Court, he explains, has “refused to extend judicial review in extradition cases, regardless of the nature of the perceived violation, absent specific direction from Congress.” Tallman dissent at 968 n. 6. Neely itself, upon which Judge Tallman rests much of his opinion, belies this assertion.
In addition to considering whether Neely could be extradited to Cuba, even though Cuba lacked many of the constitutional protections available to defendants in the United States, the Supreme Court *996also considered whether Congress had the power to pass the statute under which Neely was extradited and whether Neely’s extradition fell within the scope of that statute. These are not claims that Congress has explicitly authorized courts to review in the extradition context. Nevertheless, the Court did not refuse to review such claims. Thus, while Neely holds that it does not violate the Constitution or laws of the United States to extradite someone to a country that does not offer criminal defendants the same procedural protections as the United States, the case contains no indication that extraditions that do violate the Constitution or laws of the United States might, as Judge Tallman contends, not be subject to habeas review — indeed, it suggests the contrary.
So too does Valentine, also cited by Judge Tallman. The U.S. citizen respondents in Valentine claimed that because the relevant treaty stated that the United States was not bound to extradite its own citizens, “the President had no constitutional authority to surrender” them. Valentine, 299 U.S. at 6, 57 S.Ct. 100. The Supreme Court not only reviewed this claim, but granted habeas on that basis. Id. at 18, 57 S.Ct. 100. Similarly, here, Trinidad claims that the Secretary of State lacks the authority to surrender him. Indeed, Trinidad’s claim is arguably stronger than that of the extraditees in Valentine: CAT, as implemented by the FARR Act, does not merely state that the Executive is not required to extradite those likely to face torture; it affirmatively denies that power. I conclude that the rule of non-inquiry does not bar this claim.
As I have shown, the judicially developed rule of non-inquiry was not developed in, and does not have direct application to, judicial enforcement of obligations imposed by statute upon executive officials. The rule bars judicial examination of extraditions once it is determined that they are not contrary to the Constitution, laws, or treaties of the United States. It does not hold that we must refrain from reviewing claims that an extradition is, in fact, unlawful.
I note that this seems to be Judge Thomas’s understanding as well. Thomas cone, at 960-61. He agrees that the FARR Act limits the Executive’s authority to extradite and that courts may enforce this limitation through habeas. My disagreement with his concurrence is in how we construe the obligation the FARR Act imposes: Judge Thomas characterizes the obligation of the Secretary of State as a “duty ... to consider whether a person facing extradition from the U.S. ‘is more likely than not’ to be tortured.” Thomas cone, at 961. If this were the extent of the duty imposed by the Act, I would agree that our review of the Secretary’s compliance was limited to requiring a declaration that she had, indeed, considered whether Trinidad would be tortured upon extradition. But, as I have explained — and as the parties agree — the duty imposed upon the Secretary extends beyond simply considering whether Trinidad is more likely than not to face torture. She is required not to extradite him if there are substantial grounds to believe that he is more likely than not to face torture. Judicial review, therefore, must extend not only to determining whether the Secretary considered Trinidad’s claim that he would be tortured but to ascertaining that she complied with her obligation not to extradite where, on the available information, torture is more likely than not.
This conclusion does not resolve this FARR Act extradition case or any other. It merely establishes that substantive judicial review is not entirely precluded in this species of case and that we are writing on a clean slate in this case with regard to the *997reach of that review and the appropriate procedures that should be used.
IY. Proposed Proceedings
In approaching these questions afresh, I begin with the Supreme Court’s recognition in Boumediene that “common-law habeas corpus was, above all, an adaptable remedy. Its precise application and scope changed depending upon the circumstances.” Id. at 779, 128 S.Ct. 2229. The scope of habeas review is, in other words, not fixed. Rather, its proper application depends upon the circumstances in which it is to be applied. See id. In the circumstances of this case, I would, on habeas, apply what one might call a “rule of limited inquiry,” designed to ensure against blatant violations of the Secretary’s CAT obligations as implemented by the FARR Act.
I begin from what Boumediene identified as two “uncontroversial” features of any habeas review: (1) the detainee must have “a meaningful opportunity to demonstrate that he is being held pursuant to ‘the erroneous application or interpretation’ of relevant law”; and (2) “the habeas court must have the power to order the conditional release of an individual unlawfully detained.” Id. at 779, 128 S.Ct. 2229 (quoting St. Cyr, 588 U.S. at 302, 121 S.Ct. 2271). But, likening habeas to procedural due process, the opinion further stated that “the necessary scope of habeas review in part depends upon the rigor of any earlier proceedings.” Id. at 781, 128 S.Ct. 2229.
The underlying proceeding in Boumediene was the Combatant Status Review Tribunal (CSRT), which the Deputy Secretary of Defense had established to evaluate the enemy combatant status of those detained at Guantanamo. Id. at 733, 128 S.Ct. 2229. Habeas review of CSRT determinations, Boumediene explained, was “more urgent” than review of an ordinary judicial or even administrative proceeding, because CSRT determinations were not the result of a judicial hearing before a neutral decisionmaker. Id. at 783, 128 S.Ct. 2229. Even where “all the parties involved ... act with diligence and in good faith, there is considerable risk of error in the tribunal’s findings,” due to the “ ‘closed and accusatorial’ ” nature of the proceeding. Id. at 785, 128 S.Ct. 2229 (quoting Bismullah v. Gates, 514 F.3d 1291, 1296 (D.C.Cir.2008)). Given the consequence of error — detention for the course of an indefinite war — this “risk,” the opinion stated, is “too significant to ignore.” Id. In the CSRT context, therefore, a habeas court must have at least “some authority to assess the sufficiency of the Government’s evidence against the detainee,” as well as “the authority to admit and consider relevant exculpatory evidence that was not introduced during the earlier proceeding.” Id. at 786, 128 S.Ct. 2229.
The proceeding at issue here is quite different from the CSRTs considered by the Boumediene Court. Nevertheless, Boumediene provides some guidance. While the magistrate’s determination of a detainee’s extraditability seems to be a proceeding of the kind Boumediene held warrants minimal review, the torture determination is much more like the closed proceedings, which Boumediene held should be subject to somewhat more searching review.
In particular, the regulations governing the Secretary of State’s review of CAT claims in the extradition context are vague. It is difficult for me to determine from those regulations precisely what administrative process might be available to those claiming they are likely to be tortured if extradited. The regulations state:
(a).... In each case where allegations relating to torture are made or the issue is otherwise brought to the Depart-*998merit’s attention, appropriate policy and legal offices, review and analyze information relevant to the case in preparing a recommendation to the Secretary as to whether or not to sign the surrender warrant.
(b) Based on the resulting analysis of relevant information, the Secretary may decide to surrender the fugitive to the requesting State, to deny surrender of the fugitive, or to surrender the fugitive subject to conditions.
22 C.F.R. § 95.3.
The risk of error in the Secretary’s torture determination is likely lower than that of a CSRT tribunal — the determination is not accusatorial, nor is it likely (in general) to be directly affected by the national security concerns that underlie the CSRT determinations. It is, however, a closed process without a neutral decisionmaker. And it lacks, it would seem, any prescribed way for the detainee to present evidence or to contest that presented by the government.8 Furthermore, the consequence of error — torture—is indisputably significant, indistinguishable in terms of magnitude of harm, I would suggest, from indefinite detention. The Secretary’s determination, then, is much more like the kind of proceeding Boumediene characterized as requiring habeas review with “the means to correct errors” than one for which “habeas corpus review may be more circumscribed.” Id. at 786, 128 S.Ct. 2229.
I would hold, therefore, that in reviewing the Secretary’s determination that an individual may be extradited consistent with CAT, a habeas court must be able to inquire in some manner into the substance of the determination to “assess the sufficiency” of the Secretary’s evidence and conclusions. Id. That is, on habeas review, the court must be able to assess whether the Secretary appropriately determined that, upon extradition, torture is not more likely than not. A declaration by the Secretary (or her delegate) such as the majority requires on due process grounds, stating than that the State Department determined that a detainee may be extradited consistent with CAT may be, but is not necessarily, sufficient for that purpose. Whether it is or not depends, in my view, on what the remainder of the record shows with regard to the likelihood of torture upon extradition.
This scaled approach flows from the recognition that judicial review of the Secretary’s substantive determination should be extremely deferential. As the Government argues, the State Department is better suited than the courts to determine in the first instance the likelihood of torture and to negotiate with foreign governments to decrease that likelihood. Furthermore, the sensitivity of the foreign policy concerns implicated in extradition decisions requires that courts tread lightly in these cases. I do not take lightly the State Department’s concerns about judicial review of its torture determinations. Cf. Alperin v. Vatican Bank, 410 F.3d 532, 556 (9th Cir.2005) (stating that the State Department’s views on whether a lawsuit implicated the political question doctrine would be taken into consideration in deciding whether to exercise judicial review). I would address these concerns, however, not by refusing judicial review entirely and in all circumstances, but rather by ensuring that the procedure by which we review *999the Secretary’s torture determination is designed to take these considerations quite specifically into account. In addition, while judicial review of the Secretary’s torture determination ought generally to be available, nothing would foreclose the State Department from arguing in any particular case that under the specific circumstances of that case, we ought to decline review. Cf. Republic of Austria v. Altmann, 541 U.S. 677, 701, 124 S.Ct. 2240, 159 L.Ed.2d 1 (2004) (“[Wjhile we reject the United States’ recommendation to bar application of the FSIA to claims based on pre-enactment conduct, nothing in our holding prevents the State Department from filing statements of interest suggesting that courts decline to exercise jurisdiction in particular cases implicating foreign sovereign immunity.”).
I therefore turn to the four considerations the Government maintains militate against judicial review of the Secretary’s torture determinations. First, the Government explains that to ensure an extraditee will not be tortured, the State Department may seek assurances to that effect from a foreign government, impose conditions on extradition, and in some cases, monitor the extraditee’s treatment once in the foreign country. See Mironescu, 480 F.3d at 671-72; United States v. Kin-Hong, 110 F.3d 103, 110 (1st Cir.1997); Emami v. Dist. Court, 834 F.2d 1444, 1454 (9th Cir.1987); Sindona v. Grant, 619 F.2d 167 (2d Cir.1980); Semmelman, supra, at 1198. Arranging for these concessions, the Government argues, is a sensitive and delicate process. And their very existence may be confidential.
No doubt, judicial review of the Secretary’s torture determinations will sometimes require courts to deal with sensitive information. But we have well-developed mechanisms for dealing with such information, such as in camera review, protective orders, and procedures for reviewing classified information. See Mironescu, 480 F.3d at 673; Quinn v. Robinson, 783 F.2d 776, 788 (9th Cir.1986); Eain v. Wilkes, 641 F.2d 504, 514-15 (7th Cir.1981); see also Boumediene, 553 U.S. at 796, 128 S.Ct. 2229; Classified Information Procedures Act, Pub. L. No. 96-456, 94 Stat. 2025 (1988) (describing procedures for use of classified information in criminal proceedings); Fed.R.CivP. 5.2 (describing procedures for protective orders and filing documents under seal); Fed. R. App. P. 27-13 (describing procedures for filing sealed documents federal courts of appeals); Robert Timothy Reagan, The New “Public Court”: Classified Information in Federal Court, 53 Vill. L. Rev. 889 (2008) (describing procedures for use of classified information in recent federal civil and criminal proceedings).
I note that this is not a case in which the Government alleges that the relevant information is classified or that it is a state secret, the release of which is likely to affect national security. If the information were in either category, the Government could invoke established procedures for ensuring secrecy and, in the case of state secrets, our well-developed caselaw protecting such information (including providing for dismissal of cases in which a state secret is the subject of the proceeding or in which the case cannot go forward without information subject to the state secret evidentiary privilege). See, e.g., Mohamed v. Jeppesen Dataplan, Inc., 614 F.3d 1070 (9th Cir.2010). This case, and I expect most similar cases, involves information that implicates our diplomatic relations with other countries, but not, as far as we have been made aware, the security of our nation. While maintaining the confidentiality of this information is essential, I have no doubt of courts’ ability to do so through appropriate procedures.
*1000Second, the Government emphasizes the importance of timeliness in extradition proceedings. Delay can impede our relationships with requesting nations and may even cause some cases to become moot as statutes of limitations run out. This is a legitimate concern, one that, as in other cases in which timeliness is important, courts can address by implementing expedited procedures where requested to do so and where appropriate. Given the availability of expedition, the Government’s legitimate interest in timeliness cannot outweigh Trinidad’s right to habeas relief if his extradition would result in torture, and therefore be contrary to law. See Mironescu, 480 F.3d at 673.
Third, the Government questions the ability of courts to determine whether an individual is more likely than not to be tortured. The Government explains:
It is difficult to contemplate how judges would reliably make such a prediction, lacking any ability to communicate with the foreign government or to weigh the situation there, including the bilateral relationship with the United States, with resources and expertise comparable to those of the Department of State.
But judges routinely review such determinations. In the immigration context, courts frequently review claims that an individual, if removed, is likely to be tortured and therefore is entitled to withholding or deferral of removal under CAT and the FARR Act. See, e.g., Delgado v. Holder, 648 F.3d 1095, 1108 (9th Cir.2011) (en banc); Al-Saker v. INS, 268 F.3d 1143, 1146 (9th Cir.2001). The adjudication of these claims sometimes involves the assessment of diplomatic assurances of the kind that also may be negotiated in extraditions. See Khouzam v. Att’y Gen., 549 F.3d 235 (3d Cir.2008). There is no reason to think courts would suddenly become less competent in reviewing torture determinations simply because they were made in the context of extradition rather than immigration.
Pointing to several “obvious distinctions” between immigration and extradition, the Government argues that judicial competence to review torture determinations in immigration proceedings does not in fact indicate a similar competence in the extradition context. In particular, the Government argues that extradition treaties are negotiated “with foreign state partners meeting human rights norms,” and that “an ongoing relationship with a specific foreign state creates incentives to meet human rights commitments in extradition situations.” Given this observation, I am willing to accept that torture is less likely in the extradition than in the immigration context. But this assumption, while strongly bearing on the showing that a detainee must make in the face of a governmental declaration of FARR Act compliance such as the one the majority requires, does not justify complete displacement of all judicial authority with regard to substantive FARR Act enforcement in a habeas case.
Finally, the Government also seeks to distinguish the immigration cases based on the foreign policy implications of extradition. The Government explains that unlike immigration proceedings, extradition commences at the request of a foreign state, which commits substantial resources to the proceeding. And unlike immigration proceedings, extradition obligations are reciprocal: that is, just as the United States has agreed to extradite those who have committed crimes in certain countries, we depend on those countries to extradite individuals who have committed crimes here. If we fail to fulfill our extradition obligations, it is likely that when we request extradition, other countries will fail to fulfill their obligations.
*1001These concerns, too, are legitimate and significant. But they are largely concerns relevant to whether the United States ought to extradite an individual likely to face torture, a decision that, in ratifying CAT and passing the FARR Act, Congress has already made. They do not concern the difficulty or nature of the determination of whether torture is likely. This determination is the same regardless of whether the United States government has instigated a person’s removal or a foreign government has requested it.
In the end, the Government has not identified a single way in which the actual determination of whether a person is likely to be tortured fundamentally differs in the extradition context from that in the immigration context — let alone differs in such a way that would make courts uniformly inept in reviewing such determinations in the extradition context although uniformly competent where immigration removal is at issue.
To be sure, the State Department has experience, expertise, and diplomatic tools that courts lack. It is for this reason that I suggest that our review here, even more than our review in the immigration context, ought to be highly deferential. But the State Department’s comparative advantage in ascertaining the likelihood of torture, and in negotiating with foreign governments to ensure against torture in particular cases, does not mean that courts ought not ever review the Department’s decisions to ensure that Executive detention is in accordance with the law, which is what the majority holds. ‘Whatever power the United States Constitution envisions for the Executive in its exchanges with other nations ..., it most assuredly envisions a role for all three branches when individual liberties are at stake.” Hamdi v. Rumsfeld, 542 U.S. at 536, 124 S.Ct. 2633; see Khouzam, 549 F.3d at 250; United States v. Decker, 600 F.2d 733, 738 (9th Cir.1979) (‘We are less inclined to withhold review when individual liberty ... is implicated.”).
Given all these considerations, I would structure a habeas proceeding such as this one to minimize the burden on the State Department, to protect its legitimate interest in conducting foreign affairs, to reveal diplomatic information even to courts only when essential and when not protected by otherwise applicable doctrines, and to defer to its competence in that arena. We should therefore apply a highly deferential, limited inquiry principle to CAT claims in the extradition context, even more deferential than in the immigration context. In other words, at most, we would reverse the decision of the Secretary of State “only if the evidence is so compelling that no reasonable fact finder could have failed to find the requisite likelihood of torture.” Lanza v. Ashcroft, 389 F.3d 917, 936 (9th Cir.2004) (quoting Singh v. Ashcroft, 351 F.3d 435, 442 (9th Cir.2003)). The detainee would bear the burden of demonstrating through strong, credible, and specific evidence that torture is more likely than not, and that no reasonable factfinder could find otherwise. If, and only if, such a prima facie case is made, must the Secretary submit evidence, should she so choose and in camera where appropriate, demonstrating the basis for her determination that torture is not more likely than not.
It is premature in this ease to spell out the applicable standards in any greater detail, given that, as I stated at the outset, we do not even know whether the minimal procedural due process requirement adopted by the majority has been met. I can observe that in this case, Trinidad has made a very strong showing that his co-defendants were tortured. But the record also demonstrates that the Philippine judicial system so recognized, and that the prosecutions of the co-defen*1002dants therefore did not proceed. Under those circumstances, Trinidad would have to provide strong, credible, and specific evidence that he would be tortured if extradited even though the torture of his co-defendants made it impossible to prosecute them legally in the Philippines. I doubt the record as it stands is adequate for that purpose, and so I doubt that under my approach the government in the end would have to provide any more evidence than that the majority requires.
* * *
Habeas corpus has long been a vital mechanism for preserving the separation of powers and individual liberty. Where international interests are implicated, habeas review can implicate serious foreign affairs and separation of powers concerns. But the judiciary cannot abandon its role in preventing unlawful detention by the Executive because asserting it responsibly is not easy. Rather, where possible, we must find ways of fulfilling our obligation while addressing the executive branch’s legitimate concerns. I believe it is possible to do so in this case. The limited inquiry review I suggest above would both maintain the prerogative of the Secretary of State in conducting foreign affairs, particularly with regard to negotiations surrounding extradition, and avoid abdicating our role in preventing unlawful executive detention.
I concur in the majority’s remand to the district court for the purpose it states. If the requisite declaration is provided, I would direct the district court to conduct proceedings consistent with this opinion.
. Although Judge Tallman characterizes the government's position otherwise, it is clear to me that the government's position is that the Secretary of State may not extradite someone who is more likely than not to face torture. In addition to the statement quoted in the text above, the government, in its brief, also clearly and emphatically stated that “[t]he Government is not arguing that the Secretary of State has discretion to surrender a fugitive who more likely than not will be tortured.”
. The litigants in Pennhurst disagreed on this point. See Pennhurst, 451 U.S. at 14, 101 S.Ct. 1531.
. I note that, with the exception of Judge Kozinski, the entire en banc panel agrees that we have jurisdiction, and that the majority of us agree that the Secretary’s decision is reviewable at least to the extent of requiring an attestation of compliance with the FARR Act and CAT.
. In fact, Congress has considered and rejected legislation that would codify the rule of non-inquiry. See Jacques Semmelman, Federal Courts, the Constitution, and the Rule of Non-Inquiry in International Extradition Proceedings, 76 Cornell L. Rev. 1198, 1220-21 (1991); see also In re Extradition of Howard, 996 F.2d 1320, 1330 n. 6 (1st Cir.1993) ("The government suggests that the Constitution mandates the rule of non-inquiry. We disagree. The rule did not spring from a belief that courts, as an institution, lack either the authority or the capacity to evaluate foreign legal systems. Rather, the rule came into being as judges, attempting to interpret particular treaties, concluded that, absent a contrary indication in a specific instance, the ratification of an extradition treaty mandated non-inquiry as a matter of international comity.”).
. The statute setting forth the procedures for extradition delegates to "any Justice of the Supreme Court, circuit judge, district judge, commissioner, authorized to do so by any of the courts of the United States, or judge of a court of record of general jurisdiction of any state” the authority to charge a person with having committed an extraditable offense, issue a warrant for that person’s apprehension, and make an initial assessment of the sufficiency of the evidence against the person and certify the person’s extraditability. 18 U.S.C. § 3184. Although judicial officers are involved in this initial determination of extraditability, they are not acting in their Article III capacity — indeed, they are often not Article III judges. I will therefore refer to the judicial officers making extradition determinations in the first instance as "magistrates.”
. Several courts have suggested, though no case has yet been decided on this basis, that there may be a "humanitarian exception" to the rule of non-inquiry. See, e.g. Lopez-Smith v. Hood, 121 F.3d 1322, 1326-27 (9th. Cir.1997); Prushinowski v. Samples, 734 F.2d 1016, 1019 (4th Cir.1984); Gallina v. Fraser, 278 F.2d 77, 79 (2d Cir. 1960) ("We can imagine situations where the relator, upon extradition, would be subject to procedures or punishment so antipathetic to a federal court's *995sense of decency as to require a reexamination” of the rule of non-inquiry); see also Munaf, 553 U.S. at 702, 128 S.Ct. 2207 (distinguishing Munaf, in which the Supreme Court denied habeas from "a more extreme case in which the Executive has determined that a detainee is likely to be tortured but decides to transfer him anyway”). Because I believe that the rule of non-inquiry, as it has heretofore been developed, is not directly applicable to this case, I need not address the possibility that an exception to the rule might apply.
. The Government states that "Trinidad was given multiple opportunities to submit any material he desired to the State Department to support his claim” and that "he was further offered an opportunity to present evidence at an in-person meeting with a State Department official.” However, the Government also states that Trinidad "had no right to ... a hearing,” and, indeed, under the State Department regulations, it would seem that individuals claiming they are likely to face torture have no prescribed rights at all.