An Illinois appellate court recently imposed a $15,000 sanction against a petitioner's attorney for submitting briefs riddled with fabricated case citations and AI-generated statutory misquotations. Scott v. Illinois Human Rights Commission, 2026 IL App (1st) 251462 (July 28, 2026), adds to a growing body of authority making clear that courts will not tolerate unverified AI-generated legal authority.

The Key Points

• What changed: The First District Appellate Court sanctioned petitioner's counsel $15,000, at a rate of $1,500 per false citation or misquotation, for submitting briefs containing four false statutory quotations, one nonexistent case, and five cases that did not support the propositions for which they were cited.
• Why it matters: The court found the violations willful, referred the matter to the Attorney Registration and Disciplinary Commission (ARDC), and explicitly rejected the notion that AI hallucinations are mere "technical deficiencies."
• What to do now: Attorneys using AI drafting tools must independently verify every citation against the actual source text before filing. Reliance on a "premier corporate subscription" of an AI tool is not a defense.

Background

The petitioner dual-filed a race discrimination charge with the Department of Human Rights and the EEOC in July 2022. After the EEOC issued its determination in August 2022, the petitioner did not submit that determination to the Department within the 30-day statutory deadline, a condition the court held jurisdictional under section 7A-102(A-1)(1)(iv) of the Illinois Human Rights Act. The Commission sustained dismissal, and the petitioner appealed. Before reaching the merits, the court was confronted with a significant briefing problem.

What the Court Found

Petitioner's attorney acknowledged using ChatGPT to "synthesize complex matters" and claimed to have cross-referenced citations with LexisNexis. However, the briefs contained fabricated quotations of multiple statutes, including a fictional "full investigation" requirement attributed to a real statutory section that contained no such language, as well as citation to a case that does not exist.  Several cases were misrepresented for the propositions they were said to support.

Critically, after respondents identified the errors, the attorney did not correct them in his reply brief. He repeated the fabricated statutory language and dismissed the errors as "technical deficiencies." The court found this conduct willful.

The court was equally direct in its response: "The only acceptable standard is zero false citations." 

Practical Implications

The court's $1,500-per-misstatement rate was deliberately set higher than prior precedent, with the court explaining that increasing fines are necessary given the persistence of the problem across jurisdictions. Attorneys should note several practical points from the opinion:

• Verification must be source-level. Cross-referencing AI output against a legal research platform is insufficient if the attorney does not confirm that the cited text actually appears in the cited source.
• AI tool quality is irrelevant. The court rejected any inference that paid or "premier" AI subscriptions reduce the attorney's verification obligation.
• Silence after identification compounds the violation. Failing to correct known errors, or worse, repeating them, converts a careless mistake into willful misconduct.
• ARDC referral is now a standard consequence. The court ordered the clerk to transmit the opinion to the Illinois ARDC, consistent with practice emerging in the Fourth District and other jurisdictions.

Conclusion

Scott reflects a clear and accelerating judicial consensus: unverified AI-generated citations are sanctionable misconduct, not a byproduct of evolving technology that courts will absorb. With fines rising, ARDC referrals becoming routine, and courts explicitly rejecting "inexperience" as mitigation, the risk calculus for attorneys who skip manual verification has shifted considerably. Practitioners should watch for the application of AI usage standards, and whether the ARDC takes additional disciplinary action in response to referrals like this one.