Professor Ira Robbins, in his Essay, Against an AI Privilege “argues that—at least under current technological, social, and institutional conditions—any such privilege would be premature, unworkable, and inconsistent with the historically rooted approach to evidentiary privileges.” This approach is seen in Hon. Jed Rakoff’s Memorandum in U.S. v. Heppner that rejects the attorney-client and work product privileges for the “written exchanges that” the accused “had with a generative artificial intelligence (‘AI’) platform.”
Judge Rakoff starts by noting the high and increased usage of AI, and then asks the question to be resolved, “whether, when a user communicates with a publicly available AI platform in connection with a pending criminal investigation, are the AI user’s communications protected by attorney-client privilege or the work product doctrine?” He answers “no.”
The case that this question is being examined in is a white collar case (charges of securities fraud, wire fraud, conspiracy to commit securities fraud and wire fraud, making false statements to auditors, and falsifying corporate records.) The argument for the items being privileged relate to documents the accused had allegedly input into Claude for “the purpose of speaking to counsel to obtain legal advice.” The court going through the typical criteria for attorney-client privilege, finds it fails to meet the legal standard, noting that “Claude is not an attorney” and that “[n]o relationship exists, or could exist, between an AI user and a platform such as Claude.” The court notes that confidentiality is lacking and that the accused “did not communicate with Claude for the purpose of obtaining legal advice.” A similar analysis is seen with respect to the work-product doctrine.
I agree with Judge Rakoff’s statement that “[g]enerative artificial intelligence presents a new frontier in the ongoing dialogue between technology and the law.” But I am not as certain as to whether the attorney-client privilege or work-product privilege should not be allowed when someone uses AI. There are many questions that need to be considered as AI develops.
Although some may say that we will never see Artificial General Intelligence (AGI), it is clear that many developers are moving in this direction. Some of the questions that lawyers and the public need to consider are whether a closed or fenced AI system might present a different result? Will there be AI lawyer systems for the public to use and will that create a different scenario? Will an AI system eventually be able to establish a Kovel relationship? Should it make a difference if an attorney says to a client to work through their materials using AI? Will allowing the use of AI for those with counsel put those unable to afford counsel in a pre-indictment stage at a disadvantage? Should we be considering what is a critical stage for counsel in light of AI? There are so many more questions that need to be developed here – definitely something for scholars to start thinking about. But Judge Rakoff is correct in stating that this presents “a new frontier.”
(esp)