The University of Virginia Archival AI Protocol

The University of Virginia Archival AI Protocol (UVA AAIP) establishes a practical standard governing how artificial intelligence systems may access and use archival collections. Grounded in a core rule - no access without control - the Protocol prohibits irreversible AI model training on archival materials unless item-level provenance, meaningful attribution, and contractually enforceable institutional control can be demonstrated. The Protocol distinguishes between retrieval-based AI systems, which keep source materials under organizational control, and general-purpose model training, which absorbs knowledge into model weights irreversibly. Built on three foundational pillars - provenance and attribution, donor and community responsibilities, and institutional control - the Protocol provides a decision framework, sample contract clauses for deeds of gift and vendor agreements, minimum provenance standards for AI-generated citations, and a phased implementation plan. Designed for adoption by cultural heritage collections of any size and setting, and applicable as an AI governance framework for memory institutions broadly, the Protocol offers a consistent position from which to evaluate, negotiate, and govern AI partnerships while protecting donors, communities, and institutional trust.

Resource type: Protocol

Version 1.1, January 27, 2026. This is a living document subject to periodic review and revision.