Can documents produced by AI systems be legally privileged? Stacie Bourton, Tom Whittaker & Beata Kolodziej consider the lessons to learn from recent cases
- Two recent US decisions considered whether documents created using publicly available AI tools are protected by privilege—with contrasting outcomes.
- A judgment in England & Wales has given an indication into how the courts here will approach confidentiality where documents are AI-created.
- Factors relevant to potential future disputes about the intersection between AI and privilege, and steps organisations can take to mitigate the risk of waiving privilege in AI-generated documents.
On 10 February 2026, the US District Court for the Southern District of New York ruled that documents created by a defendant using a publicly available generative artificial intelligence (AI) tool pending a criminal investigation and sent to their lawyer were not covered by attorney-client privilege or work product privilege (United States v Heppner, No 25-cr-00503; the Heppner case).
On the same day, the US District Court for the Eastern District
.tmb-mov69x69.jpg?sfvrsn=28793f10_1)
