AI Generated Image (not the actual case documents)
Government lawyers can now attack an asylum applicant’s credibility with boilerplate asylum declarations taken from other people’s cases, without producing those people for cross-examination, under a new Board of Immigration Appeals precedent. In Matter of V-S-A-, 30 I&N Dec. 37 (BIA 2026), decided Sept. 10, 2026, the Board held that redacting personal details from the other declarations is enough to protect asylum confidentiality, and it vacated an Indian man’s asylum grant.
The ruling gives Department of Homeland Security trial attorneys a ready playbook when they suspect a story came from a template: collect similarly worded declarations from other files, black out the names and ask the judge to compare the language. Because the decision is a published precedent, it binds every immigration judge in the country, and it arrives as asylum grants in immigration court have fallen to a small fraction of decisions.
Twelve declarations, filed two days before trial
The respondent said members of a rival political party attacked him twice in India because he supported the Mann Party. He filed his declaration on Oct. 2, 2025. On Oct. 6, DHS filed 12 declarations from other Indian asylum applicants, with identifying information redacted, arguing that they shared substantial factual similarities with his account and bore “boilerplate language hallmarks.” The immigration judge gave them little weight, citing due process, privacy and the lack of cross-examination, and granted asylum on Oct. 8.
The Board rejected each of those reasons. The declarations were offered only to show repeated words and phrases, not to prove anything their authors claimed, so “cross-examination of the declarants is not required.” Redaction, the Board held, satisfies 8 C.F.R. 1208.6, the regulation that generally bars disclosing information from an asylum application to third parties.
Same words, not same facts
The more consequential correction is about method. The judge had distinguished the 12 declarations by their broad facts: different parties, different motives for the attacks, different local officials. The Board said that was the wrong comparison. Drawing on the Ninth Circuit’s decisions in Singh v. Garland (2024) and Singh v. Bondi (2025), it said judges should look for identical phrases, the same grammatical and punctuation mistakes, distinctive language, repeated narrative structure and unnecessary extraneous detail. Two accounts can involve different parties and still come from the same template.
V-S-A- clarifies rather than replaces Matter of R-K-K-, 26 I&N Dec. 658 (BIA 2015), the Board’s framework for what it calls inter-proceeding similarities. An applicant must still receive notice of the similarities, a meaningful opportunity to explain them, and a credibility ruling based on the totality of the circumstances. What the new decision removes is the argument that those safeguards include confronting the people who wrote the other statements.
The Board then added an alternative holding: even if the respondent were credible, two assaults months apart, after which he needed only medication and suffered no serious injury, did not rise to persecution. It sent the case back for the judge to reweigh credibility with the redacted declarations, decide whether he has a well-founded fear of future persecution, and rule on withholding of removal and Convention Against Torture protection, which the judge never reached. The Board expressed no view on the final outcome.
Where the pressure lands
Applicants carry the explanation. Once DHS files matching declarations, the practical burden shifts to the applicant to explain why his words resemble someone else’s. Blaming a shared preparer or translator is unlikely to be enough on its own; proof that the underlying facts are real, such as medical records, party membership documents or witness testimony, is likely to matter more. The Board took the same evidence-over-assertion approach in its October precedent on late amended asylum applications and in its ruling that judges need not accept motion to reopen declarations at face value.
Timing is the open question. Here the government’s filing came two days before the merits hearing. R-K-K- requires a meaningful chance to respond, and the headnotes in V-S-A- set no minimum notice period. Respondents’ counsel are likely to answer late filings of this kind with requests for continuances, which would add time to dockets already strained by a record appeals backlog.
Appeals will be harder to win. In Urias-Orellana v. Bondi, decided March 4, 2026, the Supreme Court unanimously held that federal appeals courts must review the agency’s determination of whether facts amount to persecution under the deferential substantial-evidence standard. If this case ends in a denial, the Board’s alternative persecution ruling would be difficult to overturn in the Ninth Circuit, where the case arose.
The decision comes as asylum grants in immigration court have dropped sharply. According to EOIR’s Asylum Decisions table, immigration judges issued 26,599 asylum grants and 82,431 denials in fiscal 2025, and 7,231 grants against 89,740 denials in fiscal 2026 through the third quarter. Based on those EOIR figures, grants fell from about 24 percent of grant-or-deny decisions to about 7.5 percent.
This article is for general information and is not legal advice.
Sources
- Matter of V-S-A-, 30 I&N Dec. 37 (BIA 2026), decision PDF
- EOIR, Agency Decisions, Volume 30
- EOIR Adjudication Statistics, Asylum Decisions (data generated July 24, 2026; FY2025 and FY2026 through the third quarter)
- Urias-Orellana v. Bondi, No. 24-777 (U.S. Mar. 4, 2026), slip opinion
- 8 C.F.R. 1208.6, disclosure to third parties