Lex Blog Briefing: AI Raises New Privilege Risks in Internal Investigations

Lex Blog has released a briefing note on AI's impact on legal privilege in internal investigations.

Professional and financial services firms are adopting generative AI (GenAI) at a rapid pace—and a new briefing note explores how this disrupts legal privilege protection during internal investigations.

• Courts are already setting precedent. In United States v. Heppner, a federal court found that communications with the AI platform Claude did not qualify for attorney-client or work product protection.
• According to a survey from the Association of Corporate Counsel, in-house counsel are highly concerned about the potential for AI use to breach privilege, citing a lack of clear internal guidelines.
• The UK’s Upper Tribunal has warned that uploading confidential material to public AI tools can amount to placing it in the public domain, risking both confidentiality and privilege status. (See legal analysis.)

Legal experts underline the stakes. Lloyd Firth states, “AI does not change the law on legal professional privilege, but it materially increases the risk of losing it.” Richard Twomey notes that sharing confidential or privileged material with AI tools can trigger underappreciated privilege questions.

With most organizations expecting GenAI to become central to their workflow within five years, and the average internal investigation spanning millions of documents, the potential for inadvertent privilege waiver is only growing.

Compliance and legal leaders overwhelmingly value digital tools but must weigh these benefits against the heightened risks AI introduces, demanding robust guidance and process updates to protect privileged information.

Yes, but: Only 18% of organizations currently measure AI ROI, complicating efforts to quantify risks and benefits.

What's next: Lex Blog's second installment on AI and privilege is forthcoming, with further guidance expected.