It is only August, and the Board of Immigration Appeals has already spent 2026 producing precedent decisions at a pace that suggests somebody at EOIR installed a button labeled MAKE IMMIGRATION LAW WORSE and then rested a brick on it.
The remarkable thing is not simply that respondents keep losing. Respondents have always lost cases before the BIA, sometimes deservedly, and bad facts do not become good facts merely because the government is unsympathetic. What is remarkable about this year’s crop is the accumulation of rules that make protection harder to obtain, government allegations easier to credit, procedural defaults more catastrophic, and Immigration Judges less able to do anything useful about any of it.
A beating during political detention is not severe enough. A government’s demonstrably ineffective efforts against organized crime may nevertheless defeat CAT acquiescence. A domestic-violence survivor’s proposed social group is simultaneously too circular and insufficiently particular. Social distinction should generally be measured across an entire country rather than in the actual community where the alleged social distinction exists. DHS may invoke an asylum cooperative agreement orally, need not necessarily produce the instrument containing its operating procedures, and may obtain pretermission without an evidentiary hearing when the record lacks sufficiently individualized evidence.
Then, on August 13, the Board decided that advance parole really does involve a “departure” after all, overruling fourteen years of its own precedent.
Apparently what immigration law lacked was more traps.
What follows is a tour through some of the most consequential and, at times, genuinely deranged BIA decisions of 2026 so far: what they hold, why the reasoning matters beyond the individual respondent, and why practitioners should be paying very close attention.
1. Matter of Delcarmen-Lara, 29 I&N Dec. 830 (BIA 2026): Congratulations on Your Government-Authorized Travel. It May Now Trigger a Ten-Year Bar.
For fourteen years, Matter of Arrabally and Yerrabelly, 25 I&N Dec. 771 (BIA 2012), stood for an enormously important proposition. A person who temporarily left the United States pursuant to advance parole did not thereby make a “departure” that triggered the unlawful-presence inadmissibility provisions of INA § 212(a)(9)(B).
There was a perfectly sensible logic behind the rule. A departure on advance parole was “qualitatively different” from an ordinary departure because the government itself authorized the temporary trip in contemplation of the person’s return. The traveler was not simply packing a suitcase, abandoning the United States, and later trying to come back. The federal government had affirmatively created a mechanism allowing that person to travel temporarily and return to continue pursuing immigration benefits.
On August 13, 2026, the Board overruled Arrabally. It now holds that leaving pursuant to advance parole is a “departure” because, stripped to its dictionary definition, to depart means to leave. The opinion actually consults Merriam-Webster for the proposition that “depart” means “to go away: leave,” which is certainly one way to dismantle fourteen years of immigration practice.
The problem is not that dictionaries are illegitimate tools of statutory construction. The problem is pretending that identifying the ordinary meaning of “departure” answers the entire interpretive question. Arrabally recognized that statutory terms operate within statutory schemes and that a government-authorized temporary trip, undertaken with advance permission to return, occupies a peculiar legal position. Delcarmen-Lara largely drains that institutional context from the inquiry and announces that leaving is leaving.
The resulting regime is absurd enough to explain itself. The government may effectively tell a noncitizen, “We grant you advance permission to leave temporarily and return,” while immigration law simultaneously warns, “By using the permission we just granted you, you may trigger a ten-year inadmissibility bar.”
That is not coherent administration. It is a side quest designed by a malicious dungeon master.
To the Board’s credit, it makes the new rule prospective, expressly recognizing the reliance interests generated by Arrabally. That softens the immediate damage. It does not make the new interpretation any less perverse.
For practitioners, the practical point is straightforward: advice about advance-parole travel that was routine under Arrabally cannot simply be recycled after August 13, 2026. A government document bearing the word “advance” does not, regrettably, mean that the government has thought through what happens next.
2. Matter of L-L-R-, 29 I&N Dec. 799 (BIA 2026): Family Reunification, but Make It a Good-Moral-Character Bar
In Matter of L-L-R-, the respondent’s involvement with a smuggler who brought his own child into the United States constituted assisting or aiding another person to enter unlawfully under INA § 212(a)(6)(E)(i). That conclusion, in turn, triggered the statutory good-moral-character bar in INA § 101(f)(3), making the respondent ineligible for cancellation of removal.
There is a textual argument for the result. The alien-smuggling provision is broad, and immigration statutes have never been especially sentimental about parents facilitating unlawful entry by their children. The perversity emerges from the statutory interaction. Conduct undertaken to reunite a parent with his own child is converted, through cross-reference, into a legal declaration that the parent cannot establish good moral character.
That phrase matters. Congress did not call § 101(f) “the miscellaneous list of statutory cancellation disqualifiers.” It chose a moral label, and immigration law then applies that label mechanically to circumstances in which moral judgment is anything but simple.
A parent can make a desperate, unlawful, even objectively bad decision for reasons rooted in family attachment, fear, poverty, or separation. The law is entitled to impose consequences on that decision. What is harder to stomach is the transformation of the conduct into a conclusive answer to an entirely different question about the person’s moral character.
Immigration law performs this trick with impressive regularity. It takes a complicated human event, passes it through several cross-referenced provisions, and eventually spits out a moral conclusion that nobody actually had to make.
Nothing says “family values” quite like converting family reunification into a character defect.
3. Matter of L-F-R-, 29 I&N Dec. 787 (BIA 2026): Welcome to Asylum as a Subscription Service
In Matter of L-F-R-, the Board held that an applicant who fails to pay the required annual asylum fee within the prescribed deadline should have the asylum application pretermitted or deemed abandoned. The decision correctly distinguishes asylum from statutory withholding and CAT protection, which are not extinguished by the annual-fee requirement.
An important caveat belongs at the outset. The BIA did not invent the annual asylum fee. Congress did. Criticism of the decision should therefore focus on what the Board actually did rather than pretending three appellate immigration judges spontaneously created Asylum Plus.
What the Board did was attach a breathtaking procedural consequence to nonpayment. A person may have a credible asylum claim. The persecutor may still exist. The applicant’s testimony may remain unchanged. Country conditions may remain unchanged. Nothing about the merits of the protection claim has necessarily deteriorated by one atom. Nevertheless, asylum itself may vanish because the annual payment was not timely made.
There are obvious governmental interests in fee collection and compliance with statutory requirements. The troubling question is why an unpaid fee should function as a merits-annihilating device rather than an ordinary financial obligation subject to collection, cure procedures, or some consequence proportionate to the violation.
The answer supplied by the decision is that the application can simply be pretermitted or treated as abandoned.
Thus asylum has finally adopted the customer-service philosophy of a cable company: your protection package has expired because we could not process your annual payment.
The only difference is that Comcast generally does not send you back to the country where you claim someone intends to kill you.
4. Matter of A-C-M-, 29 I&N Dec. 703 (BIA 2026): Why Hold a Hearing When You Can Decide There Is Not Enough Evidence to Justify One?
The Board’s asylum-cooperative-agreement decisions form one of the most disturbing clusters of the year, and Matter of A-C-M- supplies the first piece.
The Board held that when an applicant is subject to an asylum cooperative agreement and the record contains no evidence of an individualized risk of harm in the proposed third country, an Immigration Judge generally need not conduct an evidentiary hearing to determine the applicability of the safe-third-country bar.
There is an internal logic to this. Courts do not ordinarily hold evidentiary hearings merely for sport, and a party who bears a burden must eventually produce evidence capable of creating a material factual dispute.
The problem is that immigration court is not summary judgment practice in federal district court, and the respondent in A-C-M- was unrepresented. The rule creates an uncomfortable circularity: the applicant may need sufficiently individualized evidence already in the record to earn the kind of hearing at which an ordinary litigant might assume evidence could be developed and tested.
That is a formidable expectation to impose on represented respondents. Applied to a pro se asylum seeker navigating a rapidly changing safe-third-country regime, it becomes almost theatrical.
The practical message is that the applicant must arrive ready to litigate the factual danger of transfer before the tribunal necessarily needs to give her a meaningful evidentiary opportunity to litigate the factual danger of transfer.
Kafka might have rejected the setup as insufficiently subtle.
5. Matter of E-A-R-M-, 29 I&N Dec. 746 (BIA 2026): DHS Apparently May Alter the Stakes With the Litigation Equivalent of “By the Way, Ecuador”
The next installment is Matter of E-A-R-M-, where the Board held that DHS’s oral motion to pretermit the respondents’ applications for asylum and related protection provided sufficient notice of its intent to remove them to Ecuador pursuant to an asylum cooperative agreement.
The issue is not whether oral notice can ever satisfy due process. Of course it can. The problem is the institutional asymmetry exposed by the decision.
Immigration practitioners operate in a system obsessed with procedural exactitude. Respondents may suffer devastating consequences from missed filing deadlines, defective service, incorrect forms, missing translations, absent certificates, improper filing methods, late exhibits, and failures to satisfy increasingly elaborate administrative requirements. Lawyers have developed entire professional neuroses around checking whether page 247 of an exhibit packet contains the proper translation certification.
Against that backdrop, DHS may apparently announce in court that it intends to invoke an agreement that could determine whether the United States adjudicates the asylum claim at all and which third country receives the applicants, and the Board is satisfied that the oral motion supplies the required notice.
The contrast practically writes its own satire. When the respondent has a procedural obligation, the system demands Swiss-watch precision. When the government has one, “Your Honor, Ecuador” may apparently get the job done.
There may be circumstances in which oral notice genuinely provides ample time and opportunity to respond. But that should make the adequacy of the particular notice a serious functional inquiry, not an occasion to normalize informality precisely where the consequences are enormous.
6. Matter of N-E-R-S-, 29 I&N Dec. 753 (BIA 2026): You Do Not Necessarily Get the Whole Rulebook Either
If E-A-R-M- relaxes what DHS must provide by way of notice, Matter of N-E-R-S- goes after what DHS must provide by way of the governing framework.
The Board held that DHS need not provide the implementing instrument containing an ACA’s operating procedures or a complete set of operative terms and criteria in order to establish that a respondent is subject to the agreement for purposes of the safe-third-country bar.
That holding would be easier to swallow if the omitted material were administrative trivia. It was not. The agreement contemplated subsequent operating procedures containing criteria relevant to transfer. The Board nevertheless concluded that the implementing instrument was not itself part of the ACA in the sense necessary to require its production.
The procedural picture produced by the trilogy is therefore remarkable. DHS can invoke the third-country mechanism. Oral notice may suffice. The government need not necessarily place the complete operating framework before the tribunal. An evidentiary hearing may be unnecessary if the applicant has not already supplied adequate individualized evidence.
Meanwhile, the respondent is expected to figure out whether she is subject to the arrangement, identify the legally relevant exception, develop individualized evidence concerning danger in a country where she may never have lived, and do so against a governmental mechanism whose operational details may not all be before the court.
At that point, the proceeding begins to resemble less an adjudicatory process than a bureaucratic escape room in which DHS has the answer key and the applicant is told that requesting to see it demonstrates insufficient preparation.
Taken together, A-C-M-, E-A-R-M-, and N-E-R-S- reveal a striking distribution of procedural generosity. The government’s burdens become light and flexible precisely as the respondent’s become technical, anticipatory, and unforgiving.
The seesaw is perfectly balanced, provided one ignores that DHS is sitting on the ground while the asylum seeker is forty feet in the air.
7. Matter of S-E-M-Z-, 29 I&N Dec. 680 (BIA 2026): Your Family May Be Socially Distinct, but Has the Entire Republic of Honduras Been Consulted?
Particular social group doctrine was already one of immigration law’s great monuments to unnecessary abstraction. Matter of S-E-M-Z- somehow makes it worse.
The Board held that the “social distinction” element of a PSG generally must be measured on a countrywide basis, rather than from the perspective of a neighborhood or another geographically limited part of society.
The conceptual problem is profound. Human beings do not experience “society” exclusively as a nation-sized abstraction. Social meaning is created inside villages, neighborhoods, tribes, clans, religious communities, ethnic communities, professions, castes, families, and other groups that may be intensely legible to the people who inhabit them and completely irrelevant to somebody hundreds of miles away.
If every person in a particular village knows which family has been marked by a gang, the relevant persecutors know which family has been marked, and the family itself occupies a recognizable social position in that community, it is not obvious why a dentist in another province must also recognize the group before asylum law is permitted to notice it.
The countrywide approach substitutes national generality for social reality. Persecution is frequently hyperlocal. Social identification can be hyperlocal. Political power can be hyperlocal. Gang control can be hyperlocal. Tribal organization can be hyperlocal. The Board’s answer is nevertheless that social distinction ordinarily should be tested across the country as a whole.
Imagine applying that method to any other social phenomenon. Everyone in a particular town may know that two families have been feuding for generations, but apparently we cannot confidently say the families are socially distinct until the nation has been polled.
The persecutor, inconveniently, does not require Gallup’s assistance before deciding whom to target.
8. Matter of V-A-B-, 29 I&N Dec. 621 (BIA 2026): Domestic-Violence PSG Doctrine Once Again Discovers That Every Possible Formulation Is Somehow Wrong
In Matter of V-A-B-, the proposed particular social group was “married Mexican women who are unable to leave their relationship.” The Board held that the group was not cognizable because it was circularly defined and lacked particularity. It also held that a lawful marriage cannot simply be presumed from cohabitation or the existence of children.
The broader problem will be painfully familiar to anyone who has litigated domestic-violence asylum cases.
Formulate the group broadly and the government objects that it lacks particularity. Make the group more specific and it becomes circular. Define it by gender and nationality and it is too broad. Add relationship status and the formulation may become insufficiently precise. Describe the inability to leave, which is often the central social condition explaining why the abuse becomes inescapable, and the government argues that persecution has been smuggled into the definition itself.
At some point the doctrinal exercise begins to resemble Lucy holding the football for Charlie Brown. The applicant is repeatedly invited to formulate the “correct” social group, while the acceptable formulation retreats another ten yards every time she approaches it.
The decision also reflects a deeper problem with treating “unable to leave” as though it were merely a synonym for “being abused.” In many domestic-violence cases, inability to leave describes a social condition produced by legal, cultural, economic, familial, and institutional constraints. Whether that condition can satisfy PSG requirements should turn on careful analysis of the society and record, not an incantation of the word “circular.”
Immigration law instead continues searching for the Platonic domestic-violence social group that is specific enough to exist, broad enough not to be circular, socially distinct enough to count, and somehow entirely independent of the circumstances that make its members vulnerable.
Best of luck to everyone involved.
9. Matter of D-G-E-A- & N-G-G-E-, 29 I&N Dec. 570 (BIA 2026): Your Opposition to the Gang Is Not Political Unless the Gang Has Apparently Taken Political Science
In Matter of D-G-E-A- & N-G-G-E-, the Board held that opposition to criminal gangs is not, standing alone, sufficient to establish political opinion. More significantly, it described a qualifying political opinion as an actual or imputed belief or conviction concerning a discrete cause tied to a government, including a de facto government.
The first proposition is unremarkable. A person who refuses to pay a criminal because she would prefer to keep her money has not automatically expressed a political opinion. Asylum law has always required something more than a personal dispute with a criminal organization.
The difficulty lies in the Board’s effort to tether political opinion to a governmental or de facto governmental framework. In many countries, gangs and cartels do not simply commit crimes alongside a functioning state. They tax neighborhoods, regulate commerce, dictate movement, recruit children, punish cooperation with police, control territory, select local officials, corrupt security forces, enforce social rules, and decide who may live or operate a business within areas they dominate.
At some point, opposition to such an organization can become political even if the organization has neglected to issue a constitution and apply for observer status at the United Nations.
Political opinion should turn principally on the meaning attributed to the person’s conduct, including by the persecutor, within the political and social structure that actually exists. It should not depend upon whether American adjudicators can comfortably place the persecutor inside a tidy taxonomy labeled “government,” “de facto government,” or “mere criminal gang.”
People confronting an extortion demand rarely pause to determine whether their persecutor satisfies the relevant criteria for sovereign authority. The gang itself may nevertheless understand resistance as a challenge to the social order it violently administers.
Reality is often politically messier than the BIA would prefer.
10. Matter of A-H-D-, 29 I&N Dec. 642 (BIA 2026): Three Days of Political Detention and a Beating Are Apparently Not Quite Persecutory Enough
Few phrases in asylum jurisprudence have done more emotional heavy lifting than “does not rise to the level of persecution.”
In Matter of A-H-D-, the respondent was detained by police for three days after participating in a political march and beaten once during that detention. The Board held that the episode did not constitute persecution because the beating did not cause significant injury. The decision also held that a government’s general deference to tribal mechanisms for resolving tribal conflict does not itself demonstrate that the government is unable or unwilling to control persecutors within the tribe.
The cumulative facts matter. This was not an accidental shove during a street dispute. It involved state detention, political activity, confinement lasting several days, and physical violence inflicted while the respondent was in custody.
The Board nevertheless emphasizes the absence of significant injury, continuing a deeply troubling tendency to make bodily damage do work that the persecution inquiry should assign to context, coercion, state power, duration, fear, and political purpose.
Persecution is not a workers’ compensation schedule. A government does not need to fracture the correct number of ribs before detention and physical abuse acquire persecutory significance.
One is tempted to imagine an EOIR punch card on the bench: one beating recorded; four additional beatings required before the presumption of future persecution becomes available.
The tribal-protection portion of the decision raises a related concern. There are legitimate reasons a state may permit customary or tribal institutions to resolve internal disputes. But when the question is whether the state can or will protect a particular person from persecution, deference to a nonstate mechanism cannot become a talisman. What matters is whether protection is actually available and effective in the individual case.
A government may call something “tribal dispute resolution.” The person being hunted may have another name for it.
11. Matter of H-L-S-M-, 29 I&N Dec. 767 (BIA 2026): The Government Tried to Stop the Gangs, and Apparently Effort Deserves a Participation Trophy
Matter of H-L-S-M- deserves particular attention from CAT practitioners.
The applicant had cooperated with U.S. law enforcement against gang members and feared torture if removed to Mexico or Guatemala. The Immigration Judge relied on country evidence and expert testimony regarding corruption in finding CAT protection warranted. The Board reversed, emphasizing individualized considerations including the absence of past torture, the passage of time, and the vague or indirect nature of threats.
Those are legitimate considerations under the CAT framework. General country conditions cannot substitute entirely for proof that this particular applicant is more likely than not to be tortured.
The more alarming part of the decision concerns acquiescence. The Board held that the applicant failed to establish likely governmental acquiescence where the record showed that elements within the governments had taken action against criminal organizations, even though those efforts had not been effective.
That proposition threatens to turn governmental effort into a participation trophy.
The relevant inquiry cannot simply be whether some component of the state opposes organized crime in the abstract. Governments are not single sentient organisms. A federal police unit can arrest gang members while local officers accept bribes from the same gang. Prosecutors can bring cartel cases while municipal officials provide information about witnesses. A president can publicly condemn organized crime while individual security officials turn a blind eye to torture.
The existence of governmental efforts is relevant evidence. It should not become a shortcut around the harder question of whether public officials would acquiesce in the torture of this particular person.
Otherwise the inquiry begins to sound absurdly simple: Did the anti-gang program prevent the threatened torture? No. Is corruption still widespread? Yes. Are criminal organizations still capable of operating through public officials? Yes. Has somebody in government nevertheless tried to fight them?
Excellent. No acquiescence.
CAT deserves better than a gold star for governmental effort.
12. Matter of A-G-O-F-, 29 I&N Dec. 772 (BIA 2026): A Child Was Drugged and Raped, but Please Tell Us More About the Government Programs
Matter of A-G-O-F- contains the kind of facts that make the abstraction of asylum doctrine particularly difficult to stomach.
The lead respondent suffered severe abuse as a child, including being drugged and raped by two men. The Immigration Judge granted humanitarian asylum. The Board reversed.
Among its holdings, the Board concluded that “Brazilian children” and “abandoned Brazilian children” were not sufficiently particular social groups. It further held that the Immigration Judge clearly erred in finding the Brazilian government unable or unwilling to control the perpetrators, noting that one perpetrator had been arrested in connection with other crimes, that the respondent had not established why another perpetrator remained at large, and that Brazil provided services intended to protect abused children.
Those facts are certainly relevant. A functioning prosecution, effective protective services, and evidence of governmental response can defeat an unable-or-unwilling claim.
The danger lies in confusing the existence of institutions with the availability of protection.
A child-protection agency can exist on paper while a child remains completely unprotected. A perpetrator can be arrested for unrelated conduct without the government ever responding effectively to the persecution at issue. A hotline, shelter, statute, police unit, or government ministry may demonstrate institutional capacity, or it may simply demonstrate that somebody once designed a program.
Pointing to the existence of protective infrastructure while the protection failed is the asylum-law equivalent of pointing at the fire extinguisher while the building burns.
The legal question is not whether Brazil has services for abused children in the abstract. It is whether, on the evidence presented, the state was unable or unwilling to protect this child from the harm he suffered.
The decision is also a bleak illustration of asylum law’s capacity to make severe human suffering disappear behind doctrinal classification. Nobody needs a judicial opinion to determine that drugging and raping a child is horrific. The legal inquiry instead becomes whether the victim was harmed on account of membership in precisely the right social group and whether the government’s failure occurred in precisely the legally cognizable fashion.
We have managed to convert refuge into a logic puzzle.
13. Matter of Orozco Becerra & Orozco Becerra, 29 I&N Dec. 600 (BIA 2026): They Are Children, but the Statute Has a Calendar
In Matter of Orozco Becerra, the respondents included minors who failed to appear for their hearing. The Immigration Judge administratively closed the proceedings. The Board held that, once DHS establishes proper notice and removability, the Immigration Judge lacks authority to administratively close the cases and must proceed under the statutory in absentia framework, including where the respondents are children.
The Board’s statutory argument is not frivolous. INA § 240(b)(5)(A) uses mandatory language, and adjudicators cannot simply rewrite statutes because their application produces harsh facts. The Board also notes the availability of statutory mechanisms for seeking rescission of qualifying in absentia orders.
The problem is what the formal theory of notice obscures about children.
Children generally do not maintain household mailing addresses. They do not control whether their parents update EOIR after moving. They do not open legal correspondence, arrange transportation to court, select counsel, understand scheduling notices, or independently decide whether to attend a master calendar hearing. The law necessarily routes these obligations through adults and then attributes the procedural consequence back to the child.
Once formal notice is satisfied, however, the machinery moves forward. The child may receive an in absentia removal order and later attempt to unwind it through the famously simple and forgiving process known as a motion to reopen immigration proceedings.
Presumably this can be handled during recess.
The case illustrates the peculiar cruelty of treating procedural notice as though it were synonymous with meaningful capacity to comply. A six-year-old can be legally notified of a hearing without possessing anything resembling adult control over whether she appears.
The statute may constrain what an Immigration Judge can do about that problem. It does not require us to pretend the problem is sensible.
14. Matter of Shentu, 29 I&N Dec. 595 (BIA 2026): The FBI Says You May Be Dangerous. Good Luck Proving the FBI Wrong From Detention.
Finally, Matter of Shentu concerns the weight given to an FBI Letterhead Memorandum identifying a respondent as a potential national-security risk in immigration bond proceedings. The Board held that the memorandum warranted significant weight.
There is nothing inherently irrational about considering credible national-security information in a custody determination. Courts would be derelict if they categorically ignored reliable evidence suggesting that release presents a serious danger.
The difficulty is epistemic and procedural. The government possesses the investigative apparatus. The government generates the security assessment. The government introduces the assessment into a proceeding in which the detained respondent carries a substantial burden. The Board then accords the government’s assessment significant weight.
The respondent, meanwhile, must somehow rebut a national-security judgment whose underlying investigative universe may be largely inaccessible to him.
That creates a magnificent little evidentiary circle. The FBI says you may pose a national-security risk. You deny posing one. The denial carries limited weight because, naturally, a person who posed such a risk might deny it. What evidence can you produce disproving the government’s assessment? Ideally, one supposes, you would obtain the FBI’s underlying investigative material and demonstrate why the FBI misinterpreted it.
One imagines the request: “Dear Federal Bureau of Investigation, please provide the sensitive investigative information necessary for me to prove that the Federal Bureau of Investigation’s assessment of me is mistaken. I remain in detention and appreciate your prompt attention.”
The issue is not whether the FBI deserves zero credibility. The issue is what procedural mechanism exists to meaningfully test a consequential executive-branch assessment when the respondent lacks anything approaching equivalent access to the underlying information.
A system in which one party possesses the evidence, characterizes the evidence, and then receives institutional deference for its characterization demands unusually careful scrutiny.
Shentu instead gives practitioners another reason to worry about just how heavy the respondent’s burden can become once the words national security enter the room.
These Decisions Are Not Random
It would be comforting to dismiss these cases as a miscellaneous collection of fact-bound respondent losses. That reading becomes increasingly difficult as they accumulate.
Across the particular-social-group cases, the Board is demanding greater abstraction and broader social recognition while continuing to police proposed groups for excessive breadth, excessive narrowness, circularity, and insufficient particularity. The result is a doctrinal architecture in which the target keeps moving while the Board assures everyone that a valid PSG theoretically exists somewhere.
In unable-or-unwilling cases, the existence of governmental programs, arrests, services, or enforcement efforts increasingly does substantial work even when the protection afforded to the individual was obviously inadequate. In CAT cases, governmental efforts against organized crime can undermine acquiescence even when those efforts have failed to suppress the organizations whose cooperation with public officials is precisely what the applicant fears.
The ACA decisions shift in the same direction procedurally. DHS may provide oral notice. It need not necessarily produce the complete implementing instrument. The Immigration Judge may not need to hold an evidentiary hearing where the respondent has not already developed sufficient individualized evidence. The government’s side of the proceeding becomes flexible at precisely the moment the applicant’s burden becomes anticipatory and exacting.
Elsewhere, family reunification can trigger a statutory good-moral-character bar. Nonpayment of an annual fee can extinguish asylum. Minor children can be placed into the in absentia machinery once formal notice has been satisfied. A national-security assessment generated by the government’s own investigative apparatus can receive significant weight against a detained respondent expected to rebut it.
Then Delcarmen-Lara arrives to tell everyone that using government-authorized advance parole really was a statutory “departure” all along.
The common thread is not simply “the government wins.” Some of these cases involve statutory language that genuinely constrains the Board, and serious criticism requires acknowledging that fact. The more important pattern is a jurisprudence increasingly comfortable with formal governmental compliance and increasingly demanding of individualized noncitizen compliance.
When the government has done something, the existence of the governmental act itself often carries substantial weight. It authorized a program. It arrested somebody. It created child-protection services. Some officials fight gangs. The FBI issued a memorandum. DHS orally stated its position.
When the noncitizen has done something, by contrast, the inquiry becomes mercilessly exact. Was the fee paid by the deadline? Was the social group formulated correctly? Was the danger sufficiently individualized before the hearing? Was the entire country shown to perceive the group? Was the beating severe enough? Was the government’s failure to protect proven in precisely the right way? Was the political opinion sufficiently tied to governmental authority?
This is asymmetry dressed in administrative vocabulary.
Nothing dramatic is happening, according to the opinions. The group is merely insufficiently particular. The beating merely does not rise to persecution. The government’s anti-gang efforts are merely ineffective. The notice is merely oral. The operating procedures are merely unnecessary. The evidentiary hearing is merely unwarranted. The asylum application is merely abandoned. The child was merely properly notified. The FBI memorandum merely deserves significant weight. The advance-parole trip was merely a departure.
At the end of all that moderation, someone gets deported.
Welcome to Immigration Law in 2026
Immigration lawyers are accustomed to bad precedent. We are accustomed to statutes drafted like cursed crossword puzzles and standards that change with the Attorney General, the circuit, the procedural posture, and occasionally what appears to be the phase of the moon. We are accustomed to explaining to clients that words such as “admission,” “conviction,” “child,” “departure,” and apparently “persecution” do not necessarily mean what competent speakers of English might reasonably expect them to mean.
What makes 2026 different is not simply the volume of restrictive decisions. It is the consistency of the direction in which the law is being pushed and the institutional bias that direction reveals.
Again and again, ambiguity is resolved against the noncitizen. Government evidence is credited generously while evidence offered by asylum seekers is parsed for insufficiency. Government efforts that demonstrably fail to protect people are treated as proof that protection exists. DHS receives procedural flexibility while respondents are held to unforgiving deadlines, evidentiary burdens, and technical requirements. A government’s statement may be enough; an applicant’s testimony requires corroboration. An oral motion may provide sufficient notice; a missed fee may extinguish asylum. A government program may demonstrate protection even when the person standing before the court was not protected by it. An FBI assessment may receive significant weight even when the detainee has little meaningful ability to test the information beneath it.
This is not neutrality producing unfortunate outcomes by coincidence. It is a system in which skepticism is distributed asymmetrically.
The noncitizen must prove more, explain more, anticipate more, document more, and survive more procedural traps. The government, by contrast, is repeatedly afforded presumptions of regularity, institutional credibility, procedural latitude, and the benefit of abstractions that somehow always seem to point toward the same result: denial.
That matters because the Board is not merely deciding individual appeals. It is manufacturing precedent that Immigration Judges and DHS attorneys will cite thousands of times in cases far removed from the facts that produced it. A narrow loss becomes a broad rule. A factual distinction becomes a doctrinal requirement. Language describing one inadequate record becomes a checklist imposed on the next hundred asylum seekers.
The bias therefore reproduces itself.
A Board decision does not remain confined to the respondent whose name appears at the top of the page. It becomes the government’s next motion to pretermit, the Immigration Judge’s next oral ruling, the next reason an applicant supposedly failed to carry a burden that did not exist six months earlier. What begins as a published decision becomes infrastructure.
That is why precision matters more now, not less. Practitioners should read the opinions themselves rather than accept the government’s characterization of them. They should preserve contrary circuit authority, statutory arguments, and constitutional objections. They should resist every attempt to transform fact-dependent holdings into categorical rules. They should build country-condition records with painful specificity, distinguish the existence of nominal government programs from evidence of actual protection, and insist that “some officials oppose gangs” is not remotely the same proposition as “public officials will not acquiesce in this applicant’s torture.”
They should also force adjudicators to identify exactly what rule they believe a precedent establishes and exactly where the opinion says it. A remarkable amount of terrible immigration law becomes substantially worse in the retelling, particularly when the person doing the retelling begins with the assumption that the immigrant is supposed to lose.
Most importantly, practitioners must keep making records. Bias thrives when assumptions go unnamed. Put the inconsistency on the record. Put the unequal evidentiary treatment on the record. Put the government’s unsupported assertion on the record. Put the overlooked testimony on the record. Put the contradiction between the Immigration Judge’s stated standard and the actual precedent on the record.
Today’s grotesque BIA precedent may be tomorrow’s petition for review, and federal courts cannot correct what the administrative record does not preserve.
At the current pace, there is every possibility that by next Thursday the Board will publish a decision holding that being chased across the border by a death squad constitutes voluntary departure, provided DHS submits a sufficiently confident oral motion explaining why.
The opinion will presumably conclude by observing that the respondent failed to establish otherwise.