Bonilla-Espinoza v. Bondi: Sua Sponte Enforcement of Issue Exhaustion and a Demanding Proof Standard for CAT “Torture”

Court: United States Court of Appeals for the Tenth Circuit

Date: January 27, 2026

Disposition: Petition for review denied; in forma pauperis granted

1. Introduction

Bonilla-Espinoza v. Bondi involves a Salvadoran national, Jorge Alberto Bonilla-Espinoza, who sought asylum, withholding of removal, and protection under the Convention Against Torture (CAT) after multiple arrests and episodes of abuse by Salvadoran police. He alleged that he was repeatedly detained and beaten, including under El Salvador’s “state of exception,” and that his mother’s association with a human-rights organization (FESPAD) contributed to the family’s targeting.

The central issues were (i) whether Petitioner established the required nexus between harm and a protected ground for asylum/withholding (political opinion or particular social group), (ii) whether his past mistreatment and the country conditions demonstrated a likelihood of “torture” under the CAT, and (iii) whether alleged record-development failures or language/translation problems deprived him of due process.

2. Summary of the Opinion

The Tenth Circuit denied the petition for review. It held:

  • Asylum/Withholding: Substantial evidence supported the agency’s finding of no nexus between the harm and a protected ground. The court rejected the claim that police persecution was “because of” Petitioner’s anti-government political opinion.
  • Particular Social Group (“imputed gang membership”): The court refused to address the theory because it was not exhausted before the BIA. Importantly, it exercised discretion to raise exhaustion sua sponte even though the government did not press it on appeal.
  • CAT: The record did not compel a finding that Petitioner suffered past “torture,” and the agency reasonably found he did not show it was “more likely than not” he would be tortured if returned, notwithstanding troubling country conditions.
  • Due process: The IJ sufficiently developed the record, and Petitioner failed to identify prejudicial translation errors.

3. Analysis

3.1 Precedents Cited

The opinion is notably precedent-rich, using existing doctrine to resolve each claim while adding an important, published reaffirmation of appellate discretion to police administrative issue exhaustion.

A. Standards of review and scope of review

  • Dallakoti v. Holder (619 F.3d 1264 (10th Cir. 2010)): Used for the governing framework: substantial-evidence review for facts and de novo review for legal questions, and for reciting statutory standards for asylum and withholding.
  • Orellana-Recinos v. Garland (993 F.3d 851 (10th Cir. 2021)): Supports the court’s practice of limiting review to BIA grounds while consulting the IJ’s fuller reasoning where consistent.

B. Nexus and mixed-motive asylum doctrine

  • Karki v. Holder (715 F.3d 792 (10th Cir. 2013)): Invoked to clarify that “mixed-motive” analysis (“one central reason”) is unnecessary where the agency permissibly finds no protected motive at all. Here, because the agency could find the police were not motivated by political opinion (or any protected ground), it need not parse centrality.

C. Particular social group requirements (and skepticism toward perception-by-persecutor-only)

  • Rivera-Barrientos v. Holder (666 F.3d 641 (10th Cir. 2012)): Cited for the requirements that a particular social group share a “common, immutable characteristic,” and be perceived by society as discrete (the decision references the social-perception/social-distinction concept).
  • Escamilla v. Holder (459 F. App'x 776 (10th Cir. 2012)): Used as persuasive support (though unpublished) to question groups defined by a persecutor’s perception rather than society’s perception. This framed the court’s observation that “imputed gang members” may fail PSG criteria.

D. Issue exhaustion before the BIA—and appellate discretion to enforce it

  • Awuku-Asare v. Garland (991 F.3d 1123 (10th Cir. 2021)): Cited for the requirement that a petitioner must present the “same specific legal theory” to the BIA before advancing it in court.
  • Ribas v. Mukasey (545 F.3d 922 (10th Cir. 2008)): Reinforces that unexhausted issues will not be considered on the merits.
  • Santos-Zacaria v. Garland (143 S. Ct. 1103 (2023)): The Supreme Court held exhaustion rules are subject to waiver and forfeiture; the Tenth Circuit acknowledges this.
  • Miguel-Pena v. Garland (94 F.4th 1145 (10th Cir. 2024), cert. denied, 145 S. Ct. 545 (2024)): The key bridge from Santos-Zacaria to this case: even if the government forfeits an exhaustion objection, “forfeiture binds only the party[,] not the court,” and the court may raise exhaustion sua sponte. Bonilla-Espinoza applies that discretion to refuse review of an unexhausted PSG theory.

E. Agency primacy and remand principles for new theories

  • INS v. Orlando Ventura (537 U.S. 12 (2002)) and Gonzales v. Thomas (547 U.S. 183 (2006)): These cases are used to justify restraint: appellate courts generally should not decide in the first instance matters committed to agency expertise, including PSG recognition without prior agency resolution.

F. CAT standards and “torture” threshold

  • Ritonga v. Holder (633 F.3d 971 (10th Cir. 2011)): Cited for the CAT standard (“more likely than not” torture) and to reiterate the record-only review rule (declining to consider extra-record country material).
  • Singh v. Garland (No. 22-9563, 2023 WL 3337210 (10th Cir. May 10, 2023) (unpublished)), Nche v. Garland (No. 20-9621, 2021 WL 4164050 (10th Cir. Sept. 14, 2021) (unpublished)), and Bolainez-Vargas v. Garland (861 F. App'x 159 (10th Cir. 2021)): Though unpublished, these decisions are used to illustrate the “high bar” for compelling reversal on the “torture” finding and that beatings/detentions (even serious ones) may still fall short of torture as defined in the regulations.
  • Hernandez v. Garland (52 F.4th 757 (9th Cir. 2022)): Cited comparatively for a body of case law finding beatings causing scarring/hearing damage did not necessarily rise to torture.

G. Due process in immigration proceedings: record development and translation prejudice

  • Matumona v. Barr (945 F.3d 1294 (10th Cir. 2019)): The court notes it has “not explicitly recognized” an IJ’s affirmative duty to develop the record for pro se applicants, but even assuming such a duty, the IJ’s efforts were adequate and nonprejudicial.
  • Hadjimehdigholi v. INS (49 F.3d 642 (10th Cir. 1995)): Translation imperfections do not equal a due process violation absent unfair prejudice or inability to present the case.
  • Kheireddine v. Gonzales (427 F.3d 80 (1st Cir. 2005)): Cited for the requirement of “specific prejudice” tied to the asserted procedural defect.
  • Zorig v. Holder (387 F. App'x 865 (10th Cir. 2010)): Used as an example where petitioners identified translation errors through interpreter review—highlighting what was missing here: concrete error identification.

3.2 Legal Reasoning

A. Nexus failure foreclosed asylum and withholding

The court treated the nexus inquiry as the fulcrum for asylum and withholding. Even though Petitioner presented evidence suggestive of systemic abuse (including the state of exception and alleged rights violations), the agency could reasonably find the police acted for a non-protected reason: suspicion of gang involvement and attempts to extract information—conduct the IJ characterized as directed broadly at young men in Petitioner’s neighborhood.

The opinion’s reasoning emphasizes evidentiary coherence: Petitioner’s own repeated testimony about why police targeted him (“they thought he was part of [the gang] MS”; “because they want information”) undercut the protected-ground theory. The court acknowledged Petitioner’s argument that stated motives could be “pretextual,” but held the record did not compel the contrary inference required by 8 U.S.C. § 1252(b)(4)(B).

Finally, by invoking Karki v. Holder, the court underscored that mixed-motive analysis is not triggered by speculative protected motives; there must first be a protected motive supported by evidence sufficient to compel the conclusion that it existed.

B. Exhaustion and the court’s willingness to enforce it sua sponte

The most jurisprudentially significant portion of the decision concerns the proposed particular social group of “imputed gang membership.” The panel declined to reach the merits because Petitioner’s pro se notice of appeal “contained his only arguments to the agency” and did not present that PSG theory to the BIA.

Crucially, the panel recognized that under Santos-Zacaria v. Garland an exhaustion objection is waivable/forfeitable, and the government did not raise it in its merits briefing. Nonetheless, relying on Miguel-Pena v. Garland, the court exercised discretion to raise exhaustion itself and refused to decide an unexhausted claim. The rationale was practical and structural:

  • The BIA had “no occasion” to address the PSG theory or to build an explanatory record.
  • Under INS v. Orlando Ventura and Gonzales v. Thomas, appellate courts should not decide certain agency-committed issues in the first instance.
  • Even where deference might not be determinative, the court “would surely benefit” from the BIA’s analysis.

The result is a strong signal that, in the Tenth Circuit, petitioners should not assume the government’s silence will preserve unexhausted issues.

C. CAT: “torture” requires detailed proof of severe pain or suffering, not merely harsh detention and beatings

The CAT claim failed for two related reasons. First, the court accepted the agency’s view that Petitioner’s testimony lacked enough detail to establish that the past harm rose to “torture” as defined by 8 C.F.R. § 1208.18(a)(1) and limited by § 1208.18(a)(2) (“an extreme form of cruel and inhuman treatment”). The IJ repeatedly asked for specifics, and the record contained only limited descriptions (e.g., “hit me in the stomach,” head blows causing ear trauma, forced nudity and exercise, and harsh sleeping conditions).

Second, as to future risk, the panel held the agency could reasonably conclude that—even in light of troubling State Department reporting—the evidence did not compel a finding that Petitioner personally was “more likely than not” to face torture if returned. Because that dispositive standard was not met, the court declined to reach the internal relocation dispute.

D. Due process: no obligation to invent theories, and translation claims require identifiable prejudicial errors

On record development, the court rejected the contention that the IJ should have elicited testimony supporting alternative legal theories (e.g., other PSG formulations or stronger political-opinion evidence). Even assuming an affirmative duty exists under Matumona v. Barr, the IJ asked multiple follow-ups on both harm and motive. The court drew a line between clarifying facts and acting as counsel: “She had no duty to come up with potential arguments for Petitioner and then try to elicit information to support them.”

On translation/communication problems, the court applied the prejudice requirement reflected in Hadjimehdigholi v. INS and Kheireddine v. Gonzales. Petitioner did not identify specific mistranslations; confusing exchanges alone were not enough where the IJ repeatedly clarified testimony through follow-up questions.

3.3 Impact

  • Issue exhaustion after Santos-Zacaria: The decision reinforces that, even if exhaustion is not jurisdictional and even if the government does not press it, the Tenth Circuit may still enforce exhaustion sua sponte. Practically, petitioners must present PSG formulations and other “specific legal theor[ies]” to the BIA, not merely general claims of fear or unfairness.
  • CAT claims and evidentiary specificity: The opinion emphasizes that generalized accounts of beatings and harsh detention—without detailed descriptions tied to “severe pain or suffering”— may be insufficient to compel reversal on torture findings. Country conditions alone will not necessarily carry the “more likely than not” burden for the individual applicant.
  • Due process litigation discipline: Translation-based due process claims will be difficult without pinpointed errors and a clear showing of prejudice. Likewise, arguments that the IJ should have developed alternative legal theories are unlikely to succeed where the IJ asked repeated clarifying questions.

4. Complex Concepts Simplified

  • “Nexus”: For asylum/withholding, it is not enough to show harm by police; the applicant must show the harm occurred “on account of” a protected ground (race, religion, nationality, membership in a particular social group, or political opinion).
  • “Particular social group”: In this circuit, a PSG generally requires a shared immutable characteristic and that society (not just the persecutor) recognizes the group as distinct. Groups defined only by a persecutor’s suspicions can be vulnerable to rejection.
  • “Issue exhaustion”: Before asking a federal court to decide an argument, the petitioner must first present that specific argument to the BIA, giving the agency a fair chance to decide it.
  • “Waiver/forfeiture” vs. the court’s discretion: Even if the government fails to argue exhaustion (forfeiture), the court may still raise it on its own.
  • CAT “torture”: The regulations define torture as an “extreme” form of intentionally inflicted “severe” pain or suffering. Many harsh detentions and beatings may be “cruel” yet still fall short of this regulatory threshold.
  • “Substantial evidence” review: The court does not decide whether it would have ruled differently; it asks whether the record compels a contrary result.

5. Conclusion

Bonilla-Espinoza v. Bondi denies relief across asylum, withholding, and CAT, but its broader significance lies in two messages. First, asylum/withholding claims rise or fall on nexus, and courts will not infer political persecution where the record supports a non-protected motive. Second—and most importantly as a published appellate practice point—the Tenth Circuit will enforce administrative issue exhaustion even when the government does not, using its discretion to decline review of unexhausted theories such as a newly articulated particular social group. The opinion also underscores the demanding proof required to establish CAT “torture” and the need for specific, prejudicial translation error showings to sustain due process challenges.