AI Hallucination Incident

State of play.

• Sanctions are now multi-jurisdictional, escalating in dollar amount, and reaching BigLaw. The Richards Layton Leiske v. Kidd show-cause proceeding in Delaware Court of Chancery puts a major corporate firm — not a solo practitioner — at the center of a Rule 11 sanctions dispute over fabricated citations and a failed correction (→ Delaware court weighs sanctions after Richards Layton’s AI-generated brief errors, Delaware Court Orders Richards Layton & Finger to Show Cause Over AI-Hallucinated Brief).
• Illinois courts are pricing hallucinations per citation. The Illinois Appellate Court, First District imposed $1,500 per false citation — $15,000 total against attorney Mason Cole — explicitly framing the penalty as deterrence, and a second Illinois appellate panel sanctioned a separate attorney in the Starbucks bias matter (→ Illinois appeals court fines lawyer $15,000 over AI-fabricated legal citations, Illinois appeals court sanctions lawyer over AI-fabricated citations in Starbucks bias case).
• Tennessee produced the largest single-case fee award in the current cycle. A federal judge ordered Reaves Law Firm to pay Baker Donelson $45,000 under Rule 11 and circulated the sanctions order to other district judges — a deliberate deterrence broadcast (→ Tennessee Firm Sanctioned $45K for AI-Generated Fake Citations in Malpractice Suit).
• The "negligence for non-use" framing has entered the professional discourse. A University of Houston law professor has publicly posed whether failure to use AI could itself constitute negligence — inverting the current judicial consensus and signaling the next phase of the professional responsibility debate (→ Lawyer Faces Potential Negligence Claim for Failing to Use AI in Legal Work).
• For counsel advising firms on AI governance, the practical baseline is that verification failure now carries per-citation pricing, firm-level joint liability, board-level disciplinary referrals, and the first credible academic argument that non-adoption may itself become a standard-of-care issue.

Where things stand.

• Judicial consensus on attorney accountability is settled across court levels. Federal and state courts, trial and appellate, have uniformly held that unverified AI output constitutes a professional responsibility breach regardless of who in the firm ran the query (→ Delaware court weighs sanctions after Richards Layton’s AI-generated brief errors, Connecticut Supreme Court orders lawyer CLE after ChatGPT citation errors, Illinois appeals court fines lawyer $15,000 over AI-fabricated legal citations).
• Connecticut has enacted a Practice Book rule requiring independent verification of AI-generated citations. The rule applies to attorneys and pro se litigants alike, with sanctions available for violations — the most codified state-level obligation in the current corpus (→ Connecticut Supreme Court orders lawyer CLE after ChatGPT citation errors).
• Sanction severity is escalating and courts are broadcasting orders. Illinois is pricing at $1,500 per false citation; Tennessee's $45,000 award was circulated district-wide; prior Oregon rulings exceeded $100,000 — the trajectory is upward and the deterrence framing is explicit (→ Illinois appeals court fines lawyer $15,000 over AI-fabricated legal citations, Tennessee Firm Sanctioned $45K for AI-Generated Fake Citations in Malpractice Suit).
• Firm-level joint liability is now squarely in play. Delaware Court of Chancery Rule 11(c) permits joint liability against the law firm absent exceptional circumstances — the Richards Layton proceeding is the first high-profile test of that mechanism against a major firm (→ Delaware court weighs sanctions after Richards Layton’s AI-generated brief errors, Delaware Court Orders Richards Layton & Finger to Show Cause Over AI-Hallucinated Brief).
• Disciplinary referrals are running parallel to monetary sanctions. Tennessee referred Reaves Law to the Board of Professional Responsibility; Illinois referred Mason Cole; Connecticut's federal court referred Miller to the Federal Grievance Committee — sanctions and bar proceedings are now routinely paired (→ Tennessee Firm Sanctioned $45K for AI-Generated Fake Citations in Malpractice Suit, Illinois appeals court fines lawyer $15,000 over AI-fabricated legal citations, Connecticut judge fines attorney $3,500 over AI errors in taco trademark case).
• Pro se AI use has reached the trial stage. A Georgia state court proceeding went to jury trial with a self-represented litigant who used AI to prepare — extending the hallucination management problem from written filings into live courtroom proceedings (→ Georgia pro se litigant opens jury trial after using AI to help his case).
• The "negligence for non-use" argument is now in professional circulation. Academic framing has shifted: courts are simultaneously sanctioning AI misuse and the profession is beginning to debate whether competent practice will soon require AI adoption (→ Lawyer Faces Potential Negligence Claim for Failing to Use AI in Legal Work).
• ABA Formal Opinion 512 remains the governing ethics framework. It places the verification obligation on the supervising attorney; state bars including California have layered mandatory human-review and documentation requirements under Model Rule 5.3 (→ Lawyer Faces Potential Negligence Claim for Failing to Use AI in Legal Work).

Latest developments.

• Richards, Layton & Finger faces Rule 11(c) show-cause proceedings in Delaware Court of Chancery after a corrected brief left material AI-generated inaccuracies unresolved in Leiske v. Kidd; firm president Paul Heath has filed an affidavit acknowledging the errors (→ Delaware court weighs sanctions after Richards Layton’s AI-generated brief errors).
• Delaware Court of Chancery issued the initial show-cause order against Richards Layton and director Richard P. Rollo, demanding a full accounting of AI tool use, prompt entry, output integration, and the firm's written AI governance policies in effect at the time of filing (→ Delaware Court Orders Richards Layton & Finger to Show Cause Over AI-Hallucinated Brief).
• Reaves Law Firm ordered to pay Baker Donelson $45,000 under Rule 11 after fabricated citations surfaced in a malpractice suit; the court referred the firm to the Tennessee Board of Professional Responsibility and circulated the order district-wide (→ Tennessee Firm Sanctioned $45K for AI-Generated Fake Citations in Malpractice Suit).
• University of Houston law professor Renee Knake Jefferson raises the "negligence for non-use" argument — could failure to adopt AI constitute a breach of reasonable diligence — as courts simultaneously sanction AI misuse (→ Lawyer Faces Potential Negligence Claim for Failing to Use AI in Legal Work).
• Illinois Appellate Court, First District sanctions attorney Mason Cole $15,000 at $1,500 per false citation for ten fabricated authorities in a brief, explicitly framing the penalty as deterrence against "unacceptable" AI misuse (→ Illinois appeals court fines lawyer $15,000 over AI-fabricated legal citations).
• Illinois appellate court unanimously sanctions a separate attorney in the Starbucks workplace discrimination matter for AI-fabricated citations and quotations (→ Illinois appeals court sanctions lawyer over AI-fabricated citations in Starbucks bias case).
• Connecticut Supreme Court sanctions attorney Ian Gottlieb — the first Connecticut Supreme Court AI-hallucination ruling — ordering six additional CLE hours in ethics and law office management for ChatGPT-generated fabrications in two landlord-tenant appeals (→ Connecticut Supreme Court orders lawyer CLE after ChatGPT citation errors).
• Connecticut federal judge fines attorney Hilary B. Miller $3,500 and refers him to the Federal Grievance Committee after finding "subjective bad faith" in two AI-error motions in the Bartaco trademark dispute, striking both motions from the docket (→ Connecticut judge fines attorney $3,500 over AI errors in taco trademark case, Connecticut Judge Fines Lawyer $3,500 Over AI-Fabricated Quotes in Taco Trademark Case).
• Georgia state court proceeds to jury trial with an AI-assisted pro se litigant — the first documented instance of AI-assisted self-representation reaching the trial stage (→ Georgia pro se litigant opens jury trial after using AI to help his case).

Active questions and open splits.

• Will firm-level joint liability become the standard sanction structure? Delaware Court of Chancery Rule 11(c) permits it; the Richards Layton proceeding is the first major test against a BigLaw firm — the outcome will define whether firm-wide exposure, not just individual attorney sanctions, becomes the norm (→ Delaware court weighs sanctions after Richards Layton’s AI-generated brief errors, Delaware Court Orders Richards Layton & Finger to Show Cause Over AI-Hallucinated Brief).
• Is the per-citation pricing formula the emerging national benchmark? Illinois at $1,500 per false citation follows Oregon's earlier per-error approach — but no circuit has formally adopted a uniform formula, and amounts remain highly variable across jurisdictions (→ Illinois appeals court fines lawyer $15,000 over AI-fabricated legal citations).
• Does "subjective bad faith" attach even when the attorney apologizes and commits to remediation? The Connecticut federal court found bad faith despite Miller's stated remorse and remedial promises — but whether that standard applies uniformly or only where the court finds the verification failure was willful is unresolved (→ Connecticut judge fines attorney $3,500 over AI errors in taco trademark case, Connecticut Judge Fines Lawyer $3,500 Over AI-Fabricated Quotes in Taco Trademark Case).
• What AI governance documentation satisfies a court's inquiry? Delaware Chancery demanded the firm's written AI policies, who entered prompts, how output was integrated, and whether the signing attorney personally reviewed cited opinions — no court has yet defined what documentation is sufficient to avoid or mitigate sanctions (→ Delaware Court Orders Richards Layton & Finger to Show Cause Over AI-Hallucinated Brief).
• How will courts manage AI use by pro se litigants at trial? Disclosure and verification obligations for represented parties are developing; for pro se litigants in live proceedings — examination, argument, real-time courtroom work — no framework yet exists (→ Georgia pro se litigant opens jury trial after using AI to help his case).
• Will "negligence for non-use" become a cognizable malpractice theory? The academic argument is now in circulation; courts are simultaneously sanctioning AI misuse and signaling AI adoption may become a professional standard — the window for firms to develop protocols before the standard shifts is narrowing (→ Lawyer Faces Potential Negligence Claim for Failing to Use AI in Legal Work).
• What disclosure obligations attach to AI use in filings? Some courts have standing orders requiring AI disclosure; Connecticut has codified a Practice Book rule; no uniform federal rule or circuit-wide standard has emerged (→ Connecticut Supreme Court orders lawyer CLE after ChatGPT citation errors, Lawyer Faces Potential Negligence Claim for Failing to Use AI in Legal Work).

What to watch.

• Vice Chancellor Will's ruling in the Richards Layton Leiske v. Kidd sanctions proceeding — the first BigLaw firm-level Rule 11(c) joint liability determination over AI hallucinations, which will set the template for firm governance obligations and remediation standards (→ Delaware court weighs sanctions after Richards Layton’s AI-generated brief errors, Delaware Court Orders Richards Layton & Finger to Show Cause Over AI-Hallucinated Brief).
• Whether the Illinois ARDC pursues formal disciplinary proceedings against Mason Cole following the First District's referral — which would be the most significant bar discipline action tied to AI hallucinations to date (→ Illinois appeals court fines lawyer $15,000 over AI-fabricated legal citations).
• Whether the Tennessee Board of Professional Responsibility takes action against Reaves Law Firm, and whether other district judges in the Western District of Tennessee begin citing the circulated sanctions order in their own AI-related rulings (→ Tennessee Firm Sanctioned $45K for AI-Generated Fake Citations in Malpractice Suit).
• Whether any state bar — California, New York, or Texas — issues formal guidance on the "negligence for non-use" question, which would mark the first official recognition that AI adoption is a component of competent practice (→ Lawyer Faces Potential Negligence Claim for Failing to Use AI in Legal Work).
• Whether Georgia courts issue formal guidance on AI use by pro se litigants in live proceedings following the jury trial, and whether other states follow Connecticut's Practice Book verification rule (→ Georgia pro se litigant opens jury trial after using AI to help his case, Connecticut Supreme Court orders lawyer CLE after ChatGPT citation errors).