- within Technology, Transport, Media, Telecoms, IT and Entertainment topic(s)
- with Inhouse Counsel
- with readers working within the Law Firm industries
The decision contains important lessons, but the analysis may not be identical under English law.
In what appears to be the first decision addressing the privilege status of communications with an AI chatbot, the Southern District of New York has held that written exchanges between a criminal defendant and a publicly available Generative AI platform (Anthropic's Claude tool) were not protected by either attorney-client privilege or the work product doctrine: US v Heppner No 25 Cr 503 (SDNY).
Attorney-client privilege is similar to legal advice privilege under English law. It protects confidential communications between client and attorney for the purpose of obtaining or providing legal advice. The court in this case found that documents recording the defendant's communications with Claude did not meet this test for a number of reasons:
- Claude is not a lawyer and "discussion of legal issues between two non-attorneys is not protected".
- The defendant had no reasonable expectation of confidentiality in his communications with Claude.
- While this aspect was seen as a closer call, the court was not persuaded that the defendant communicated with Claude for the purpose of obtaining legal advice, commenting that the question was whether he intended to obtain legal advice from Claude, and that Claude disclaims providing legal advice.
As for the work product doctrine, which is similar (but not identical) to litigation privilege under English law, the court found that this doctrine did not apply because – even assuming the documents were created in anticipation of litigation – they were not prepared by or at the direction of counsel. The court declined to follow, and explicitly disagreed with, a 2021 SDNY decision that concluded, in relevant part, that the work product doctrine is not limited to materials prepared by or at the direction of an attorney.
The decision contains important lessons, not least as to the importance of using private AI tools with strict confidentiality protections. However, it is not clear that the same conclusion would necessarily be reached in the English law context, applying the principles discussed in our recent article, "Navigating legal privilege issues when using AI". In particular, under English law, litigation privilege protects materials created for the dominant purpose of obtaining legal advice, information or evidence relating to the conduct of litigation which is reasonably in prospect. There is no additional requirement that they were created by or at the direction of a lawyer. As the Court of Appeal noted in Al Sadeq v Dechert LLP [2024] EWCA Civ 28, "litigation privilege is also enjoyed by a person acting without a lawyer in relation to actual or contemplated litigation".
Potentially, therefore – subject to the question of confidentiality – an English court might have held that litigation privilege applied in this case, if satisfied that the documents were prepared for the dominant purpose of obtaining legal advice, information or evidence in relation to anticipated litigation.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
[View Source]
