At the Public Sector InfoGov Summit 2026, focused on “Advancing Information Governance Across the Public Sector,” one message came through consistently: AI in government is not just a technology question but actually an information governance conversation. From the White House Office of Records Management and the National Archives to the Department of the Interior and Veterans Affairs, the focus was not on hype but instead on readiness, responsibility and results.
Government runs on records. And before AI can deliver value, those records must be structured, managed and trusted. Philip Droege, director of the White House Office of Records, gave a keynote on managing White House records that was a powerful reminder that governance at the highest levels of government is built on discipline, documentation and defensibility. Whether under the Presidential Records Act or federal records mandates, accountability depends on clarity: what a record is, where it lives, who owns it and how it transitions.
That same principle applies to AI.
John Montel, associate CIO from the Department of the Interior, reinforced this directly: Before introducing AI, agencies must prepare their information environment. Physical records, electronic content and data must be organized.
Ownership must be clear. Retention must be consistent. Metadata must be reliable.
Records Management Lays AI Groundwork in Government
Without that groundwork, AI amplifies inconsistency rather than value. This is where AI-powered content management becomes meaningful. AI can automatically classify documents, extract metadata, identify sensitive information and summarize complex files. But those capabilities only create impact when they operate inside a governance framework with retention schedules, access controls, audit trails and human oversight built in.
The National Archives provided a compelling example. Acting Chief Innovation Officer Jill Reilly described how NARA is leveraging AI to enhance metadata, improve semantic search and expand access to historic collections including the 1950 Census and Revolutionary War Pension files. But the key to success was not the algorithm. It was the “ground truth” datasets and strong metadata foundation beneath it. AI was layered on top of trusted records, not substituted for them.
That is the model government leaders should follow.
Dr. Moya Hill, chief FOIA/privacy/records officer with the Office of Acquisition, Logistics, and Construction, led a discussion on FOIA at the Department of Veterans Affairs which further underscored the point. Transparency is not optional. AI can assist in identifying responsive records, standardizing redactions and accelerating response times. But defensibility remains paramount with explainable outputs and auditable decisions. Governance programs must align FOIA, privacy and records management rather than treating them as separate disciplines. The conversation also turned toward agentic AI, systems capable of coordinating actions within defined rules.
AMTRAK Senior Director of Data, AI and Automation John Chiofee’s closing keynote on the “governance fabric” tied information, data and AI together. Governance cannot sit in a silo. It must travel with the information across systems, departments and workflows. AI will increasingly monitor compliance thresholds, surface anomalies and recommend actions—but within guardrails defined by policy.
That evolution mirrors what we see in AI-empowered content management today.
- Phase one is intelligent extraction structuring information automatically as it enters the system.
- Phase two is human-plus-AI collaboration accelerating review, classification and routing while preserving oversight.
- Phase three moves toward intelligent orchestration where AI agents help coordinate workflows, identify risk and support decisions under defined governance policies.
This evolution is built into Laserfiche, a platform grounded in decades of public sector experience. Records management capabilities support full lifecycle tracking, granular security, automated retention and audit readiness aligned with standards like DoD 5015.2. AI capabilities such as Smart Fields, intelligent classification and workflow automation operate within that governance framework.
The takeaway from the summit was clear: AI is not replacing records management. It is elevating it.
Government leaders should not ask, “How do we deploy AI?” The better question is, “Is our information environment ready?” Agencies that invest now in structured metadata, lifecycle governance and cross-functional collaboration between IT, records and business leaders will be positioned to move from managing records to activating intelligence.