The New York City Bar Association released new guidance Monday on attorney AI use, telling lawyers their “default practice” should be not to use AI tools to record, transcribe or summarize conversations — unless they have a particularly compelling reason to do so in a certain situation and have obtained everyone’s permission.
There are ethical issues and risks that arise when attorneys use AI to record, transcribe and summarize conversations – whether the conversation be with co-counsel, opposing parties, clients, potential clients investigators or people expected to testify at a trial or hearing, like witnesses or experts, the City Bar says.
“AI systems that record, transcribe, and summarize conversations can be useful tools, but attorneys must be mindful of the ethical issues that arise when AI is used for this purpose,” the City Bar states. “In some contexts, the risk involved in recording the conversation will outweigh the value of doing so.”
Essentially, the Bar’s guidance boils down to: while having recordings of what people say may be helpful, it can ultimately come back to bite you: For example, recording of a prospective client you don’t end up working with accessible within your firm could create the risk of information-sharing and a conflict of interest, and a recording or AI summary of a witness recording could become something an attorney is required to turn over during discovery, risking jeopardizing confidentiality, creating an opportunity for the other side to pull unfavorable evidence out of context and possibly placing protected work product at risk.
And, especially if you’re recording with AI, it’s important to ensure that recordings and summaries of conversations are not accessible beyond attorneys and others involved in a client’s representation, in order to preserve attorney-client privilege.
“Whenever attorneys contemplate recording conversations, it is important to understand how AI tools record, transcribe, and summarize conversations, and how this information is stored and shared, because issues of confidentiality and privilege are implicated,” the City Bar advises. “The recording of conversations may add obligations to review transcripts and summaries of those conversations for accuracy, thereby potentially adding to costs/fees.”
The City Bar says that, if attorneys feel strongly that recording a conversation with AI is particularly necessary in a certain situation, they can — as long as clear consent is obtained from all parties — but lawyers should really try to avoid it.
“Because the recording and preservation of conversations … may jeopardize confidentiality and privilege, and because questions will often arise about the means by which the employee or agent obtained consent, the risks of recording the conversations typically outweigh the benefits, and the default practice should be that these conversations will not be recorded,” the City Bar states.
There’s also various security issues lawyers should be aware of if they’re sending trial preparation or other confidential legal documents over a video call that’s being recorded. Confidentiality should always be assured, and attorneys should make sure that transcripts aren’t able to be reviewed by larger pools of people or fed into larger AI systems than should be.
The guidance comes as the legal profession continues to grapple with how and whether to use AI in its practice, where it’s necessary, where it’s helpful and where it hurts. Attorneys have been yelled at and held in contempt by judges for filing AI-generated documents that cite fake, or AI “hallucinated” cases, and public and nonprofit legal service providers have said they’ve found it can be a tool that their staff can leverage to help them serve more people than they’d ever be able to without it.
Columbia Law School last week released a new policy telling its students they were allowed to use generative in their studies, but strictly to help them understand legal concepts and conduct research – never to write, form their legal analysis or take exams.
“As AI transforms the study and practice of law, the challenge before us is not whether to engage with AI, but how to do so in a way that advances the Law School’s academic and educational purpose: teaching our students how to think critically and rigorously, and helping them develop the judgment, discipline, and technological fluency that modern legal practice increasingly demands,” the Law School’s Dean Daniel Abebe wrote in a statement.