As the public sector races to adopt artificial intelligence, the “regulatory paradox” has reached a fever pitch. Agencies are no longer just exploring the frontier (they are feeding their records into AI engines).

For records managers and archivists serving as the “Territorial Marshals,” the challenge has shifted. You are no longer just protecting static boxes or servers (you are managing the fuel for AI-powered workflows). This session focuses on how to ensure your agency’s transition to AI is built on a foundation of compliant, secure, and well-managed information.

• Defensible Data Minimization: Ensure only necessary records are accessible to AI (reducing “garbage in, garbage out” risk).
• Modernizing Retention for AI: Update “need-to-keep” policies so tools don’t ingest (or hallucinate from) expired or obsolete records.
• Advocating for Information Privacy: Speak the language of privacy to IT and vendors so AI architectures respect sensitive citizen data.
• Validating the Output: Define the records manager’s role in human-in-the-loop systems to keep AI-generated content accurate, authentic, and trustworthy.

The AI landscape will keep shifting, but your role as the arbiter of trust is more critical than ever. Join us as we workshop practical governance strategies that move agencies from unregulated innovation to secure, records-compliant transformation.