On Friday, August 4th, 2026, the Connecticut Supreme Court addressed issues arising from the use of generative AI in legal document preparation in the case of Tov Realty LLC v. Suarez. Attorney Gottlieb initially conducted legal research and verified citations using LexisNexis and Shepard's Citations Service. However, after pasting drafts into ChatGPT to enhance efficiency and quality, the AI introduced new or altered case citations without Gottlieb’s knowledge. These AI-generated hallucinations made the final court filings inaccurate and misleading.
Despite these citation errors, the substantive legal propositions in the filings remained accurate and well established. Consequently, neither Attorney Gottlieb nor his colleagues re-verified the altered citations before submission.
As a result, Attorney Gottlieb is required to complete six additional hours of continuing legal education focused on ethics and law office management, including three hours specifically on the use of generative AI. Furthermore, both Gottlieb and his law firm must each donate $1,000 to the CT Bar Institute, Inc. to support legal ethics and AI-related education.
The court’s decision underscores the importance of thorough citation verification after AI-assisted editing. It advises legal professionals to either re-check citations and proofread after AI edits or to use AI solely for editing suggestions that are manually reviewed and applied.
Sources
- Reason, Eugene Volokh, ChatGPT Editing Adding Hallucinations to Otherwise Accurate Material?
Loading comments.