The courts have applied traditional Federal Rules of Civil Procedure 26 relevance and proportionality standards to AI-generated materials without carving out exemptions for privilege or work product. Emerging proposals like Rule of Evidence 707 address AI evidence reliability, but no specific legislation has yet codified these principles. The scope of what constitutes discoverable AI materials remains subject to case-by-case judicial interpretation.
Attorneys should implement immediate confidentiality protocols around AI tool use. Inputting privileged communications, client information, or work product into third-party platforms—even inadvertently—now carries significant privilege waiver risk. The volume of discoverable ESI in litigation involving AI workflows has expanded substantially, creating both production burdens and exposure in civil, criminal, regulatory, and corporate contexts. Firms should establish clear policies restricting what information can be fed into AI systems and consider whether proprietary or enterprise AI solutions offer better confidentiality assurances than consumer platforms.