The New York State Bar Association has explicitly framed AI delegation and supervision as a core ethics issue. Legal teams are already using generative AI for document review, research, summarization, and drafting, with Thomson Reuters reporting that many practitioners view it as a significant productivity tool. But recent sanctions involving AI-generated fictitious case citations have exposed the risks of over-reliance on automated systems without meaningful human review. Bar associations, courts, and vendors are now issuing guidance on supervision, disclosure, and verification of AI-assisted work—signaling that the profession is moving from experimentation to operational deployment.
The boundary between safe and unsafe delegation remains unsettled. Lawyer ethics rules require competence, confidentiality, and verification of output, with particular emphasis on independent human judgment for tasks involving legal reasoning or client confidentiality. Which specific tasks can be delegated to AI without lawyer review, and which cannot, has not been clearly defined. Attorneys should expect continued guidance from bar associations and courts on this question, and should treat any AI-generated work product as requiring verification before filing or relying on it in client matters.