Statutory Precedent Survives Loper Bright: New York Child Sexual Abuse Remains an INA Aggravated Felony

1. Introduction

This decision addresses an important consequence of the Supreme Court’s rejection of administrative deference: whether Loper Bright Enterprises v. Raimondo, 603 U.S. 369 (2024), permits a circuit panel to disregard earlier statutory interpretations that relied on Chevron U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837 (1984). The Third Circuit answers no. Ending Chevron deference does not, by itself, erase the precedential force of decisions previously rendered under that framework.

The immediate setting was immigration law. Sofiqul Islam, a Bangladeshi national and lawful permanent resident, challenged the determination that his New York conviction for second-degree sexual abuse constituted “sexual abuse of a minor,” an aggravated felony under 8 U.S.C. § 1101(a)(43)(A). He also challenged the denial of protection under the Convention Against Torture (CAT).

The court rejected three asserted differences between the New York offense and the federal generic offense: the range of prohibited sexual conduct, the absence of an age-gap requirement, and the required mental state. Its age-differential analysis resolved a question of first impression for the circuit. The court separately upheld the denial of CAT protection, while expressing caution about the Board of Immigration Appeals’ (BIA) use of administrative notice to characterize changing country conditions.

2. Background and Procedural History

Political activity and immigration

Islam joined the Bangladesh Nationalist Party (BNP) as a teenager, attended meetings, helped organize rallies, distributed leaflets, and participated in community outreach. At three rallies in 1998, members of the rival Awami League attacked attendees with incendiary devices, firearms, knives, and physical violence. Islam suffered minor cuts while fleeing but was not otherwise personally harmed, arrested, or detained.

He entered the United States as a lawful permanent resident in 2009, joining his United States citizen wife. They have two children who are also United States citizens.

Conviction and removal proceedings

In 2022, Islam pleaded guilty to misdemeanor second-degree sexual abuse under New York Penal Law (NYPL) § 130.60(2), which prohibits subjecting a person under fourteen to “sexual contact.” He received a sentence of 364 days in jail. After his release, the Department of Homeland Security initiated removal proceedings.

The Immigration Judge (IJ) sustained the aggravated-felony charge, relying on Debique v. Garland, 58 F.4th 676 (2d Cir. 2023), and the Third Circuit’s nonprecedential decision in Reyes v. Att’y Gen., No. 22-1493, 2023 WL 2184831 (3d Cir. Feb. 23, 2023). Both treated the same New York provision as a categorical match for sexual abuse of a minor.

The aggravated-felony charge was the only sustained basis for removal. Although the IJ’s oral decision mistakenly suggested that a child-abuse charge had also been sustained, the BIA recognized the error. The court therefore did not decide whether the conviction independently supported removal as child abuse, neglect, or abandonment.

Claim for CAT protection

Islam maintained that, if returned to Bangladesh, he would resume BNP activity and face politically motivated violence. He also feared nighttime curfews imposed in July 2024 with shoot-on-sight orders, asserting that he would lack shelter.

The IJ found Islam credible but concluded that his evidence did not establish that torture was more likely than not. The BIA affirmed. It also took administrative notice that the Awami League government had fallen in August 2024 and that an interim government led by Muhammad Yunis had assumed power, characterizing country conditions as greatly improved.

3. Summary of the Decision

• Existing statutory precedent remained binding. Loper Bright Enterprises v. Raimondo did not authorize the panel to disregard earlier decisions merely because they had deferred to the BIA.
• The New York offense was not overbroad as to sexual conduct. Its prohibition on sexually motivated touching fell within the federal generic offense, guided principally by 18 U.S.C. § 3509(a)(8).
• No age differential was required. The absence of an age-gap element in NYPL § 130.60(2) did not create a categorical mismatch.
• The required mental states matched. New York’s requirement that touching occur “for the purpose of gratifying sexual desire” required knowing or intentional touching, not merely reckless or negligent contact.
• CAT protection was properly denied. Substantial evidence supported the finding that Islam had not established a probability of torture.
• Any questionable country-conditions inference was harmless. The IJ’s denial remained supported without relying on the BIA’s assertion that conditions had greatly improved.

Disposition: The court dismissed the petition insofar as it challenged removability and denied it insofar as it challenged the denial of CAT protection.

4. Governing Framework: Jurisdiction and the Categorical Approach

Different review rules for removal and CAT protection

Under 8 U.S.C. § 1252(a)(2)(C), courts generally lack jurisdiction to review a final removal order against a noncitizen convicted of an aggravated felony. Section 1252(a)(2)(D), however, preserves review of colorable constitutional claims and questions of law.

Pareja v. Att’y Gen., 615 F.3d 180 (3d Cir. 2010), supplied the colorability framework. Valansi v. Ashcroft, 278 F.3d 203 (3d Cir. 2002), established that whether a conviction is an aggravated felony is reviewed de novo: it is a legal question that also determines the court’s jurisdiction.

CAT review follows a different path. Under Nasrallah v. Barr, 590 U.S. 573 (2020), the aggravated-felony restriction does not foreclose review of factual challenges to a CAT denial. Those findings receive substantial-evidence review and remain conclusive unless the record would compel any reasonable adjudicator to reach the opposite conclusion.

Because the BIA adopted particular aspects of the IJ’s analysis, the court reviewed both decisions under Oliva-Ramos v. Att’y Gen., 694 F.3d 259 (3d Cir. 2012).

Comparing offenses, not reconstructing events

The court used the categorical approach described in Moncrieffe v. Holder, 569 U.S. 184 (2013), and Shular v. United States, 589 U.S. 154 (2020). It first identified the federal generic offense and then compared that definition with the elements necessary for conviction under New York law.

Mathis v. United States, 579 U.S. 500 (2016), supplied the central instruction: disregard the particular facts of the defendant’s conduct and compare the legal elements. Thus, the inquiry was not what Islam actually did, but whether the New York provision criminalized conduct outside the federal category.

New York’s misdemeanor classification and Islam’s sentence did not resolve that inquiry. “Aggravated felony” is a federal statutory designation; the sexual-abuse-of-a-minor category does not depend on whether state law labels the offense a felony.

5. Analysis of the Court’s Reasoning and Precedents

A. Loper Bright changed interpretive methodology, not the binding status of every earlier decision

The controlling Third Circuit precedents

In Restrepo v. Attorney General, 617 F.3d 787 (3d Cir. 2010), the Third Circuit deferred to the BIA’s interpretation of “sexual abuse of a minor.” The underlying administrative precedent, Matter of Rodriguez-Rodriguez, 22 I. & N. Dec. 991 (BIA 1999), used 18 U.S.C. § 3509(a)(8) as a guide rather than treating the narrower federal criminal provision, 18 U.S.C. § 2243(a), as controlling.

That choice mattered. Section 3509(a)(8), located in a statute protecting child victims and witnesses, includes molestation and other forms of sexual exploitation. Section 2243(a), read with its associated definitions, covers a narrower set of sexual acts and includes an age differential.

Cabeda v. Att’y Gen., 971 F.3d 165 (3d Cir. 2020), reaffirmed Restrepo v. Attorney General after the Supreme Court’s decision in Esquivel-Quintana v. Sessions, 581 U.S. 385 (2017). The Supreme Court there independently interpreted the generic offense in the context of age-based statutory rape, without relying on Chevron. Although the Cabeda v. Att’y Gen. majority acknowledged methodological tension with the earlier deference-based approach, it found no sufficient basis to abandon the earlier substantive interpretation.

Judge Krause had disagreed with aspects of that reasoning in her separate opinion in Cabeda v. Att’y Gen.. Her authorship of the present opinion underscores an institutional distinction: a judge’s disagreement with an existing rule does not free a later panel from following it.

Why the Supreme Court’s new methodology did not displace those cases

Islam argued that abolishing Chevron removed the foundation of both circuit decisions. The court instead emphasized Loper Bright Enterprises v. Raimondo’s express preservation of statutory stare decisis. The Supreme Court stated that earlier holdings validating agency action remained subject to ordinary precedent rules and that reliance on Chevron alone was not a special justification for overruling them.

The court rejected Islam’s narrower interpretation—that only the particular agency action and outcome remained protected, while the adopted statutory interpretation could be discarded. Statutory stare decisis ordinarily protects legal rules, not merely the result between the original parties. The discussion in Bastias v. Att’y Gen., 158 F.4th 1188 (11th Cir. 2025), particularly Judge Newsom’s concurrence, supported that distinction.

Virgin Islands v. Martinez, 620 F.3d 321 (3d Cir. 2010), supplied the panel-precedent rule: earlier precedential panel decisions remain binding absent en banc reconsideration or an intervening Supreme Court decision that conflicts with them. Here, the intervening Supreme Court decision expressly sought to preserve existing holdings.

Alternative support for retaining the interpretation

The court also explained that the earlier interpretation did not rest exclusively on administrative deference. Restrepo v. Attorney General considered ordinary meaning, statutory construction, existing circuit law, and decisions from other circuits, including decisions not applying Chevron.

Drawing on Abdulla v. Att’y Gen., 153 F.4th 342 (3d Cir. 2025), the court identified no independent legal basis for departure that had not already been considered and rejected. This reinforced the result, although the principal ground remained the binding force of precedent.

The broader circuit landscape

The decision therefore establishes continuity without claiming that every consequence of Loper Bright has been resolved. Questions involving later agency changes, conflicting interpretations, or independent grounds for reconsideration remain distinct.

B. New York’s prohibition on sexual contact fits within the generic offense

Having retained § 3509(a)(8) as the “primary guide,” the court considered Islam’s argument that New York law nevertheless covered excessively broad conduct, including kissing or brief touching of a neck, navel, or leg.

The decisive analogy: Pennsylvania indecent contact

Cadapan v. Attorney General, 749 F.3d 157 (3d Cir. 2014), addressed a Pennsylvania statute defining indecent contact in language nearly identical to New York’s definition of sexual contact: touching sexual or other intimate parts for sexual arousal or gratification.

That decision held that the prohibited conduct fell within the ordinary meanings of “molestation” and “sexual exploitation” in § 3509(a)(8). The present panel found no meaningful distinction between the two statutory definitions. Covered conduct need not involve penetration, exposed genital contact, or a narrowly specified anatomical area.

The limiting feature is important: New York does not criminalize every touch of a broadly understood intimate area. The touching must occur for the purpose of gratifying sexual desire. The court assessed the combined statutory requirements, not anatomical location in isolation.

State cases confirmed the statutes’ comparable reach

• Commonwealth v. Crenshaw, 306 A.3d 383 (Pa. Super. Ct. 2023), quoting Commonwealth v. Gamby, 283 A.3d 298 (Pa. 2022), explained that intimate parts can include personal and private body areas commonly associated with sexual relations or intimacy.
• Commonwealth v. Fisher, 47 A.3d 155 (Pa. Super. Ct. 2012), illustrated that Pennsylvania indecent contact can include touching the legs, shoulders, neck, or back.
• Rodriguez v. Barr, 975 F.3d 188 (2d Cir. 2020), quoting People v. Morbelli, 544 N.Y.S.2d 442 (N.Y. Crim. Ct. 1989), showed that New York similarly considers the location, manner, and circumstances of touching.
• People v. Graydon, 492 N.Y.S.2d 903 (N.Y. Crim. Ct. 1985), and People v. Belfrom, 475 N.Y.S.2d 978 (N.Y. Sup. Ct. 1984), supplied examples involving a child’s leg and compelled touching of the defendant’s navel.

These authorities did not independently define the federal offense. They demonstrated that the New York statute was materially comparable to the Pennsylvania provision already examined in binding circuit precedent.

Dictum and agency-rationale objections

Islam characterized the relevant discussion in Cadapan v. Attorney General as nonbinding dictum. Applying United States v. Mallory, 765 F.3d 373 (3d Cir. 2014), the court disagreed: the interpretation of molestation and sexual exploitation formed an essential analytical basis for the earlier holding.

The court also rejected an objection under Sec. & Exch. Comm’n v. Chenery Corp., 332 U.S. 194 (1947), which ordinarily prevents judicial affirmance of agency action on a rationale the agency did not invoke. The BIA had found the conviction to fall within the aggravated-felony category and had relied on decisions embracing molestation and sexual exploitation. The panel was explaining that rationale rather than substituting an unrelated ground.

Debique v. Garland directly supported that understanding. Reyes v. Att’y Gen. agreed with it, but the court expressly recognized that the unpublished Third Circuit decision was nonbinding.

C. The absence of an age-gap element does not create overbreadth

The court first corrected the state-law comparison

The court agreed with Islam that NYPL § 130.60(2) contains no age-differential element. It specifies the victim’s age, but not the perpetrator’s. Neighboring provisions expressly specify perpetrator ages, reinforcing the significance of that omission.

The IJ had relied on New York’s limitations on criminal responsibility for young defendants to infer an age gap. The court rejected that approach. General infancy rules are not necessarily elements of the particular offense.

Grijalva Martinez v. Att’y Gen., 978 F.3d 860 (3d Cir. 2020), supported focusing on offense elements rather than affirmative defenses. Descamps v. United States, 570 U.S. 254 (2013), identified elements as facts requiring unanimous proof beyond a reasonable doubt. Stubbs v. Att’y Gen., 452 F.3d 251 (3d Cir. 2006), reinforced the restriction to the statutory offense definition. New York’s model jury instructions also omitted any perpetrator-age element.

The result illustrates an important feature of appellate review: the court affirmed the categorical match while rejecting part of the IJ’s explanation for it.

Text, historical statutes, and federal structure

Esquivel-Quintana v. Sessions had left the age-differential question open. Using its analytical framework, the court examined statutory language, related federal law, dictionaries, and state criminal codes.

The phrase “sexual abuse of a minor” focuses on the victim’s age. It does not expressly require an older perpetrator. Dictionary evidence likewise supplied no age-gap requirement.

The court’s historical survey found that, when Congress added the category in 1996:

• Twenty-eight jurisdictions had at least some relevant child-sexual-abuse provisions without an age differential.
• Among twenty-nine jurisdictions with provisions addressing victims fourteen or younger, fifteen required no age gap in the statutory text.

These figures undermined the claim that a four-year differential formed part of the generic offense. They were contextual evidence of the offense’s established legal meaning, not a mechanical majority-vote test.

Taylor v. United States, 495 U.S. 575 (1990), supplied a structural analogy: courts should hesitate to adopt a purported generic definition so narrow that it excludes much of the state-law conduct Congress apparently intended to capture. Similarly, Esquivel-Quintana v. Sessions had declined to import § 2243(a) wholesale into the immigration category.

Other circuits and the distinction from adolescent statutory rape

• United States v. Escalante, 933 F.3d 395 (5th Cir. 2019), rejected an age differential for the generic offense in the Sentencing Guidelines context. It was persuasive, rather than directly controlling immigration authority.
• United States v. Rodriguez, 711 F.3d 541 (5th Cir. 2013), supplied dictionary-based support, although the opinion acknowledged its abrogation on other grounds by Esquivel-Quintana v. Sessions.
• Velasco-Giron v. Holder, 773 F.3d 774 (7th Cir. 2014), supported consulting related sentencing analysis in the immigration setting. Its discussion did not independently settle every age-related boundary of the generic offense.
• Estrada-Espinoza v. Mukasey, 546 F.3d 1147 (9th Cir. 2008), had required a four-year gap. However, United States v. Thompson, 127 F.4th 1204 (9th Cir. 2025), questioned that approach after the Supreme Court’s intervening reasoning.
• United States v. Medina-Villa, 567 F.3d 507 (9th Cir. 2009), treated the earlier Ninth Circuit rule as addressing statutory rape, while Quintero-Cisneros v. Sessions, 891 F.3d 1197 (9th Cir. 2018), described sexual conduct involving children under fourteen as inherently abusive.
• Garcia Pinach v. Bondi rejected an age-gap objection to the same New York provision, emphasizing the victim’s being under fourteen.
• Garcia-Urbano v. Sessions, 890 F.3d 726 (8th Cir. 2018), and Bedolla-Zarate v. Sessions, 892 F.3d 1137 (10th Cir. 2018), did not definitively require an age gap. They concluded that any necessary differential would not exceed the gap already present in their cases.

Islam relied on Leger v. Attorney General, 101 F.4th 1295 (11th Cir. 2024), which required at least a one-year differential for statutory rape offenses involving consensual sexual activity among adolescents aged twelve to fifteen. The Third Circuit distinguished that setting from the sexual-contact provision before it. A survey combining penetrative statutory rape laws with sexual-contact statutes did not establish a controlling generic rule for this case.

The Model Penal Code also supported Islam’s position to some extent because several of its offenses contain age differentials. Under Pugin v. Garland, 599 U.S. 600 (2023), such materials are relevant interpretive sources. They are not dispositive, however, and did not outweigh the other evidence.

Scope of the holding: The court rejected an age-differential requirement for the generic offense in resolving this sexual-contact conviction involving a victim under fourteen. Its reasoning should not be treated as silently eliminating the distinct questions surrounding consensual adolescent statutory rape.

D. The sexual-gratification requirement makes the touching knowing or intentional

Under Cabeda v. Att’y Gen., the federal generic offense requires at least knowing conduct concerning the sexual act. Islam argued that New York required a sexual purpose but allowed the actual touching to be reckless or negligent.

The court rejected that separation. NYPL § 130.00(3) defines sexual contact as touching intimate parts “for the purpose of gratifying sexual desire.” Read naturally, the purpose requirement attaches to the touching itself.

Contextual interpretation and analogous statutes

United States v. Williams, 553 U.S. 285 (2008), supplied the principle that surrounding words give statutory terms more precise meaning.

Borden v. United States, 593 U.S. 420 (2021), distinguished purposeful or knowing force from reckless force by examining the targeted meaning of the full statutory phrase. United States v. Henderson, 80 F.4th 207 (3d Cir. 2023), similarly interpreted a robbery provision’s threat language as directed toward an intended target.

The panel applied that reasoning by analogy. In context, New York’s touching language describes directed conduct toward another person’s intimate parts, not accidental contact with a mere recipient. The analysis did not depend on a universal rule that the word “touching,” standing alone, always establishes intent.

The importance of New York’s own interpretive law

People v. Smith, 591 N.E.2d 1132 (N.Y. 1992), was particularly significant. New York’s highest court interpreted a similarly structured robbery definition requiring force “for the purpose of” producing specified consequences. It held that the language necessarily required an intent to carry out the specified purpose.

The Third Circuit respected that state-law reasoning under Jackson v. Virginia, 443 U.S. 307 (1979), and Mullaney v. Wilbur, 421 U.S. 684 (1975), which recognize state courts’ authority to define the substantive elements of state crimes.

The court distinguished the Pennsylvania provision previously considered in Cabeda v. Att’y Gen.. That provision lacked an equivalent affirmative sexual-gratification purpose requirement, making resort to a general mens rea gap-filler appropriate there but unnecessary here.

Islam’s hypothetical accidental subway touching therefore failed: moving near someone for sexual gratification does not make later unintended contact a touching performed for that purpose. Garcia Pinach v. Bondi independently supported the resulting categorical match.

Important limitation: This analysis concerned awareness or intent regarding the touching. It did not establish a separate requirement that the perpetrator know the victim’s age.

6. CAT Protection: Credibility, Individual Risk, and Administrative Notice

A. Credible testimony did not establish the necessary probability

Under Myrie v. Att’y Gen., 855 F.3d 509 (3d Cir. 2017), Islam had to establish that torture by, or with the acquiescence of, a government official was more likely than not. Predictions about whether future events would occur were factual findings.

The IJ’s credibility finding meant that Islam’s account was accepted as truthful. It did not compel acceptance of his prediction of future torture. Credibility and evidentiary sufficiency are separate questions.

Applying Herrow v. Att’y Gen., 93 F.4th 107 (3d Cir. 2024), the court asked whether the evidence compelled a conclusion contrary to the agency’s—not whether another adjudicator could reasonably have granted protection.

B. The IJ considered the material evidence

Green v. Att’y Gen., 694 F.3d 503 (3d Cir. 2012), requires consideration of all evidence relevant to possible future torture. The IJ discussed attacks on BNP rallies, arrests and targeting of party members and leaders, political repression, curfews, and Islam’s own past and anticipated activities.

The court distinguished Herrow v. Att’y Gen. and Kang v. Att’y Gen., 611 F.3d 157 (3d Cir. 2010), where material evidence of widespread torture or mistreatment had been ignored or mischaracterized. Here, the agency considered the evidence but found it insufficient.

Two points supported that finding. Islam testified that he could stay with family, undermining his asserted inability to shelter during curfews. He also did not establish either a sufficiently widespread pattern of torture placing ordinary BNP supporters at the requisite risk or a prominent political role exposing him to particular targeting.

The court acknowledged inconsistent language in the IJ’s discussion of whether Islam would resume political activity. Nevertheless, accepting that he would remain a supporter and might volunteer did not fill the evidentiary gap. Hernandez Garmendia v. Att’y Gen., 28 F.4th 476 (3d Cir. 2022), supported requiring specific evidence connecting the applicant to the predicted danger.

The decision does not make party leadership a legal prerequisite for CAT relief. Nor does it exclude proof based on widespread group-based torture. It holds that the evidence here did not establish the required probability for someone in Islam’s anticipated position.

C. Noticing a political transition differs from finding improved safety

Under 8 C.F.R. § 1003.1(d)(3)(iv), the BIA may take administrative notice of commonly known facts, current events, and official documents, but generally may not undertake appellate factfinding.

Zubeda v. Ashcroft, 333 F.3d 463 (3d Cir. 2003), recognized administrative notice of commonly acknowledged facts and certain extra-record materials. The opinion noted that the case had been abrogated on other grounds by Auguste v. Ridge, 395 F.3d 123 (3d Cir. 2005).

The change of government and the interim leader’s Nobel Peace Prize were appropriately noticeable public facts. The further statement that conditions had “greatly improved” was different: it was an evaluative conclusion, unsupported by a citation to official country reports or comparable materials.

The court expressed concern but did not need to decide that the statement required reversal. Any error was harmless because substantial evidence supported the IJ’s denial under the country conditions existing at the hearing. The judgment therefore does not authorize the BIA to infer safety merely from regime change or a leader’s international reputation.

7. Impact and Limits of the Decision

Administrative law: continuity after a methodological break

The principal impact extends beyond immigration. Litigants cannot secure reconsideration of settled statutory holdings merely by showing that an earlier court used Chevron. A successful challenge must identify a legally sufficient basis beyond the discontinued methodology, while respecting the institutional limits on panel reconsideration.

At the same time, the opinion does not resolve how precedent operates when an agency later changes its interpretation. Its discussion of other circuits acknowledges that preserving precedent and avoiding permanent entrenchment of an agency’s last approved position may generate further disputes.

Immigration law: three common overbreadth arguments become harder

For NYPL § 130.60(2), the decision forecloses the particular arguments that sexually motivated touching is too broad, that the generic offense necessarily requires a four-year age gap, or that New York permits merely reckless touching. It also reinforces the use of closely comparable state statutes when applying existing categorical precedent.

The decision does not establish that every state sexual offense involving a minor qualifies. Different statutes may contain materially different conduct requirements, victim-age thresholds, or mental states.

CAT litigation: connect country conditions to the applicant

General political violence remains relevant, but applicants must connect that evidence to a probability of torture in their circumstances. Conversely, adjudicators must consider the full record and may not substitute an unsupported inference of improved safety for a supported assessment.

A notable jurisdictional formulation

After extensively analyzing Islam’s statutory arguments, the court concluded that none presented a colorable legal claim permitting review of the final aggravated-felony removal order and dismissed that portion of the petition.

This sequence deserves attention. Whether a legal argument fails and whether it is sufficiently colorable to support jurisdiction are analytically distinct questions. The opinion ultimately treated these particular challenges as insufficient to cross the statutory threshold. Its disposition should not be generalized into a rule that every unsuccessful aggravated-felony argument necessarily defeats jurisdiction.

8. Complex Concepts Simplified

Chevron deference

The former framework under which courts sometimes accepted an agency’s reasonable interpretation of an ambiguous statute. Loper Bright ended that framework.

Statutory stare decisis

The principle that an established judicial interpretation of legislation retains precedential force, even when interpretive methods later change.

Panel precedent

A later three-judge panel ordinarily must follow an earlier precedential circuit decision rather than overrule it.

Categorical approach

A comparison of legal offense elements, not an examination of what the defendant actually did.

Federal generic offense

The nationally applicable legal definition against which a state conviction is compared.

Categorical mismatch

A situation in which the state offense permits conviction for conduct outside the federal generic definition.

Mens rea

The required mental state, such as intention, knowledge, recklessness, or negligence.

Dictum

An observation unnecessary to the decision, unlike reasoning essential to its holding.

Substantial-evidence review

A deferential standard requiring the agency’s factual findings to stand unless the evidence compels the opposite conclusion.

Administrative notice and harmless error

Administrative notice permits recognition of certain established public facts without ordinary proof. An error is harmless when it does not affect the outcome.

9. Conclusion

Sofiqul Islam v. Attorney General United States of America separates the end of administrative deference from the survival of judicial precedent. Loper Bright requires independent statutory interpretation going forward; it does not automatically reopen every interpretation adopted during the Chevron era.

Applying that principle, the Third Circuit held that New York’s second-degree sexual-abuse offense matched the INA category despite the absence of an age-gap element. Its sexual-gratification requirement also supplied the necessary knowing or intentional touching. The court separately upheld the CAT denial because the record did not compel a finding that Islam would probably be tortured, while cautioning against unsupported conclusions drawn from political change.