Building the Data Foundation for AI at Scale and the Future of Agentic AI in Government

Federal agencies are entering a decisive phase in their adoption of artificial intelligence. The early wave of experimentation, characterized by pilots, proofs of concept, and limited deployments, has largely validated the potential of AI. The challenge now is fundamentally different. Agencies must determine how to scale these capabilities across complex, mission-critical environments while maintaining trust, security, and operational integrity. During a recent ATARC roundtable, leaders from across government and industry gathered to discuss how to design systems for a machine-driven future. This report outlines the key takeaways from the conversation.