AI Privilege and Legal Advice: What Managing Partners Need to Know

AI Privilege and Legal Advice: What Managing Partners Need to Know

According to an article by Mona Ebert, Phillip Kelly, Rachel Tookey, and Crystal Sawyers of DLA Piper, legal advice prepared with the assistance of generative AI may attract privilege in certain circumstances, but only within the boundaries of existing privilege doctrine. There is no recognized concept of standalone “AI privilege,” and no test case or legislative reform has yet clarified the issue. Instead, courts are likely to rely on established principles governing who gives legal advice and under what conditions it is protected.

The authors explain that advice produced directly by a generative AI system and delivered to a non-lawyer cannot be privileged, because the system itself is not a lawyer. However, privilege may be preserved when a qualified lawyer, whether in-house or in private practice, uses generative AI as a tool under their direction and supervision. In this framework, AI is analogized to a subordinate, such as a trainee solicitor or paralegal. Where a lawyer carefully reviews, adopts, and authorizes AI-assisted work product, the advice may be treated as the lawyer’s own and therefore capable of attracting privilege.

The article emphasizes that responsible use is essential. Public AI systems, particularly free tools with broad access to user inputs, are not appropriate for generating legal advice. Instead, organizations should rely on bespoke AI systems with contractual safeguards, ensure lawyer oversight, and consider documenting AI usage to demonstrate supervision. Internal policies and training are also critical to prevent inadvertent waiver.

AI can be integrated into legal workflows without forfeiting privilege, but only with disciplined governance, clear supervision, and an understanding that privilege remains grounded in the lawyer’s role, not the technology.