Background
Attorney Ian G. Gottlieb and GLG Law, LLC, filed applications for public interest appeals and subsequent briefs in TOV Realty, LLC v. Suarez and a companion case. Gottlieb initially researched the matters through LexisNexis and verified the citations in his rough drafts, but then used ChatGPT to improve the drafts’ organization and writing. ChatGPT added or altered citations, and neither Gottlieb nor the firm’s other reviewing lawyers rechecked them before filing.
After amici alerted the court and counsel to apparent hallucinated citations, Gottlieb filed errata sheets and corrected briefs. The court then ordered Gottlieb and the firm to show cause. Gottlieb admitted that the filings contained approximately seven erroneous, unverified citations and that his conduct violated the duty of competence under Rule 1.1 of the Connecticut Rules of Professional Conduct.
The Court’s Holding
In a per curiam order, the Connecticut Supreme Court agreed that filing court documents containing erroneous and unverified AI-generated citations violated Rule 1.1. The court found no intent to deceive, but emphasized that lawyers remain personally responsible for the accuracy of their filings and that negligent reliance on generative AI threatens the integrity of the adversarial process. The firm shared responsibility because it lacked policies and procedures governing responsible AI use.
The court ordered Gottlieb to complete six additional hours of continuing legal education in ethics and law-office management, including three hours concerning generative AI. It also ordered Gottlieb and the firm each to donate $1,000 to the CT Bar Institute, Inc., and to report compliance within six months. Gottlieb must satisfy any reciprocal-discipline obligations in other jurisdictions. The order will be posted publicly and sent to the Statewide Grievance Committee as a courtesy copy, not as a referral for further discipline.
Key Takeaways
- A lawyer violates the duty of competence by filing erroneous, unverified citations produced or altered by generative AI.
- Prior verification of a draft is insufficient when an AI tool later edits the document; citations and authorities must be independently checked after AI use.
- Law firms share responsibility for implementing policies and review procedures addressing generative AI, including accuracy, confidentiality, privilege, and work-product risks.
Why It Matters
This is the Connecticut Supreme Court’s first confrontation with hallucinated citations caused by a lawyer’s use of generative AI. The order makes clear that lack of deceptive intent does not excuse an attorney’s failure to verify AI-modified filings.
The decision also reinforces Connecticut’s newly adopted trial and appellate rules requiring independent verification of AI-generated citations, authorities, and evidence. Lawyers may use generative AI, but they remain accountable for every representation submitted to a court.