Dec v. Mullin: 7th Circuit Admonishes Immigration Lawyer for AI-Generated Fake Citations — But Imposes No Financial Sanction | Advocate Prakhar

Case at a Glance

Background

Danuta Dec, a Polish national, was approved for a US visa in 2009 based on a petition filed by her US citizen sister. While waiting for the visa, she was unlawfully present in the US for over a year — triggering a ten-year bar on admissibility under 8 U.S.C. § 1182. Her mother, a lawful permanent resident, died in May 2020. Dec then applied for a waiver of inadmissibility, arguing that her mother’s death triggered a statutory provision allowing continued adjudication of related applications notwithstanding the death of a qualifying relative. USCIS denied the waiver, ruling that Dec’s sister — not her mother — was the qualifying relative for her visa petition. The district court dismissed her challenge for lack of subject-matter jurisdiction.

The AI Issue

Dec’s attorney filed an opening brief at the Seventh Circuit that contained non-existent citations and a false quotation “seemingly generated by artificial intelligence.” Chief Judge Brennan, writing for the panel, addressed this directly: “This case presents a second issue. The petitioner’s opening brief included non-existent citations and a false quotation, seemingly generated by artificial intelligence. On these specific circumstances, we admonish petitioner’s attorney but impose no other sanction.” The court did not explain in detail why it chose admonishment rather than a financial penalty — but the phrasing “on these specific circumstances” suggests the court weighed factors such as whether the AI misconduct affected the outcome or caused significant harm to the opposing party.

What the Court Decided

• The district court lacked subject-matter jurisdiction over Dec’s immigration waiver challenge — Congress explicitly precluded judicial review of all agency decisions on waivers, including eligibility determinations [jurisdiction dismissed].
• Dec could not demonstrate eligibility under 8 U.S.C. § 1154(l) because her sister, not her mother, was the qualifying relative — the statute’s “surviving relative” provision did not apply [merits dismissed].
• The opening brief contained non-existent citations and a false quotation seemingly generated by AI [AI hallucination finding at circuit level].
• The attorney is formally admonished — a permanent public record in the court’s judgment [admonishment without financial penalty].
• No further sanction was imposed given the specific circumstances of this case [context-sensitive response].

“The petitioner’s opening brief included non-existent citations and a false quotation, seemingly generated by artificial intelligence. On these specific circumstances, we admonish petitioner’s attorney but impose no other sanction.”

— Chief Judge Brennan, Dec v. Mullin, 7th Circuit, March 30, 2026

The India Angle

Indian Law Equivalent

Immigration and visa matters in India are governed by the Foreigners Act, 1946 and the Passport (Entry into India) Act, 1920. Appeals against FRRO and Ministry of Home Affairs decisions lie to the High Court under Article 226 of the Constitution — there is no direct equivalent of the US immigration court/Board of Immigration Appeals system. However, the AI hallucination issue is equally relevant to Indian visa and immigration lawyers: if false citations are submitted in a High Court writ petition challenging an OCI card cancellation or visa refusal, the consequences under the Contempt of Courts Act, 1971 and Bar Council Rules are identical.

Bar Council Rules

Bar Council of India Rule 22 (no false statements) is directly engaged. Indian immigration lawyers increasingly use AI to navigate the complex web of FRRO notifications, visa category rules, and bilateral agreements — and must be especially careful because much of this material changes frequently, creating exactly the kind of data-currency gap that leads to AI hallucinations.

Practical Advice for Indian Advocates

• Immigration law changes rapidly — visa categories, fee structures, and eligibility criteria are updated by circular notifications that AI training data may not include. Always verify current rules from official FRRO/MHA portals, not AI summaries.
• A formal admonishment from a court — even without a financial penalty — is a permanent public record that can affect an advocate’s professional standing; treat it as seriously as a fine.
• For High Court writ petitions on immigration matters, every statutory provision cited must be checked against the current consolidated version — AI tools often cite provisions that have been amended or renumbered by subsequent legislation.

Quick Takeaways

• A circuit court admonishment without a fine is still a serious public record of misconduct.
• Immigration law’s rapid change makes AI-generated citations especially risky in visa matters.
• Jurisdictional defects often doom cases before hallucinated citations can even be reached — but the AI problem gets noticed anyway.

Deep Dive: Why the Seventh Circuit’s “Admonishment Only” Response Is More Significant Than It Appears

At first glance, the Dec v. Mullin outcome looks lenient compared to the $2,000 fine in Coleman, the $7,032 in Brownfield, or the $2,500 in Chaney. But the Seventh Circuit’s decision to admonish rather than fine requires careful interpretation, not dismissal. The court’s phrase “on these specific circumstances” is doing a lot of work. The circumstances the court weighed almost certainly include: the immigration case was dismissed on jurisdiction — the fake citations were in briefs that did not ultimately affect the outcome. The petitioner (not just the attorney) was already suffering a significant consequence — she lost her waiver challenge and remains subject to a ten-year re-entry bar. Imposing a financial sanction on her attorney in this context might effectively be punishing the client by potentially making her lawyer’s services more expensive in future proceedings.

But the admonishment is not nothing. It is published in a reported circuit court decision. It will appear in any future search for this attorney’s name in Westlaw, Lexis, or PACER. Future courts will know that this attorney has had a finding of AI hallucination in circuit court briefing. If the attorney appears before the Seventh Circuit again with similar issues, the prior admonishment becomes the baseline for a much more severe response — exactly as Kachouroff’s prior sanction in Coomer escalated when he repeated the conduct.

The immigration law context is worth examining specifically. Visa and immigration matters are among the most AI-hazardous practice areas for several reasons. First, the statutory scheme is labyrinthine — the Immigration and Nationality Act has dozens of interacting provisions, and small details (whether the qualifying relative is a parent or a sibling, as in Dec’s case) are dispositive. Second, immigration rules change constantly through regulatory updates, policy manuals, and agency guidance — creating a large gap between what AI training data contains and what the current rule actually is. Third, immigration clients are often vulnerable: facing deportation, family separation, or a ten-year re-entry bar. The stakes of AI-generated errors in this context are not just professional but human. Indian lawyers appearing in High Court writ petitions challenging visa refusals, OCI card cancellations, or FRRO show-cause notices face the same combination of legal complexity, rapid regulatory change, and human consequences that makes unverified AI research particularly dangerous in immigration work.