The Issue
Federal and state courts across the nation have begun imposing severe sanctions and disciplinary measures against attorneys who cite nonexistent cases and fabricated legal authorities generated by artificial intelligence. In June 2026, the trend has accelerated sharply. The 6th Circuit Court of Appeals imposed $30,000 in sanctions against two attorneys for more than two dozen fake citations; the 5th Circuit imposed $2,500 sanctions against a single attorney; and judges at the district court level have referred attorneys to disciplinary counsel for what judges describe as the most egregious Rule 11 violations they have encountered. Simultaneously, courts have begun admonishing opposing counsel who fail to identify and report fake citations in their opponents’ briefs. For criminal defense practitioners, this dual liability creates significant ethical and financial exposure. The Florida Supreme Court’s new AI rules, effective June 15, 2026, provide guidance but also set a higher bar for attorney responsibility and oversight.
Background: From Curiosity to Crisis
Generative AI tools like LexisNexis CoCounsel and vLex have become ubiquitous in law practice. These tools promise to accelerate legal research, drafting, and brief preparation. However, a critical problem emerged: large language models occasionally generate plausible-sounding but entirely fictional case citations, statutes, and legal authorities. This phenomenon, known as hallucination, results from how these models function. Throughout 2025 and into mid-2026, courts have grown increasingly hostile toward such conduct. The escalation reflects judicial frustration with two distinct failures: attorneys using AI tools without independent verification, and opposing counsel failing to catch and report fake citations.
Strategic Implications for Criminal Defense
Criminal defense practitioners face distinct risks. First, using AI for brief writing, motion drafting, or research without thorough verification exposes the attorney to sanctions, discipline, malpractice liability, and client harm. Sanctions orders can include monetary fines ($2,500 to $30,000 in recent cases), fee shifts, contempt findings, and referral to state bar disciplinary counsel. For a criminal defense firm, such consequences damage reputation and create credibility problems with judges. Second, defense counsel now face an affirmative obligation to verify opposing counsel’s citations. Courts have stated that it is troubling for counsel not to conduct cite-checks on opponent briefs. Third, AI-generated errors in criminal procedure contexts carry heightened stakes. An erroneous citation in a sentencing memorandum, post-conviction motion, or appellate brief could undermine the defendant’s entire argument. Courts may view such errors with greater skepticism in criminal cases because the defendant’s liberty interest is at stake.
How Does This Apply in Louisiana?
Louisiana courts have not yet issued comprehensive AI ethics rules comparable to Florida’s June 2026 amendments. However, Louisiana Rules of Professional Conduct Rule 1.1 (competence) and Rule 8.4 (misconduct) apply directly to AI misuse. Under La. R.P.C. Rule 1.1, attorneys must provide competent representation, which includes understanding the tools they use and their limitations. An attorney who uses AI for legal research without verification breaches the competence requirement. La. R.P.C. Rule 8.4(c) prohibits conduct involving dishonesty, fraud, deceit, or misrepresentation. Submitting a brief containing fabricated citations violates this rule. Practitioners should implement the following safeguards: (1) establish a verification protocol for AI-generated citations before any brief is filed; (2) verify all citations independently using established legal databases; (3) document the verification process in the case file; (4) disclose AI use to clients where material; (5) conduct cite-checks of opposing counsel briefs and report suspicious citations to the court if fictitious; and (6) maintain written AI policies for the firm, including supervision protocols and training.
Frequently Asked Questions
If I use AI to draft a motion and verify all citations before filing, am I protected from sanctions?
Verification greatly reduces risk but does not eliminate it. Courts recognize good-faith efforts to verify and may be more lenient if an error slips through despite reasonable precautions. Document your verification process thoroughly.
Do I have an obligation to report fake citations in my opponent’s brief?
Courts increasingly suggest yes. While courts have not imposed sanctions solely for failing to report, judges have stated that counsel should point out nonexistent citations to the court. Conservative practice is to report fake citations immediately upon discovery.
Can I be disbarred for AI citation errors?
Disbarment is unlikely for a single error, but repeated violations or a pattern of dishonesty could escalate to disbarment. More common consequences are public reprimand, suspension, or probation. The Oklahoma Supreme Court’s public reprimand against attorney Matthew Reeves is an example of a mid-range sanction.
What is the difference between AI hallucinations in civil versus criminal cases?
Courts may apply heightened scrutiny in criminal cases because the defendant’s liberty interest is at stake. Defense counsel should expect courts to be less forgiving of AI errors in criminal contexts.
Should I disclose my use of AI to clients?
Material disclosure is prudent. La. R.P.C. Rule 1.4 requires that clients be kept informed about material aspects of representation. Many firms now include AI use in engagement letters.
The era of casual AI use in legal practice has ended. Courts have spoken loudly through sanctions, discipline, and public criticism. Criminal defense counsel must implement rigorous verification protocols, train staff on AI risks, maintain written policies, and hold themselves and opposing counsel to a high standard of citation accuracy. Compliance is a core component of competent and ethical representation.