Agencies urged to build AI structures, not just plans

A May 2026 commentary argued that federal agencies will not be able to deploy artificial intelligence at scale by relying on strategy documents alone, and instead must build durable internal structures that can carry AI systems from pilots into sustained operations. The author presented the core problem as organizational rather than purely technical, with leadership choices and management design determining whether policy goals become working systems.

The piece said agencies face a dual requirement set by the White House AI Action Plan: maintain stable, secure services while also modernizing quickly. In the author’s view, federal managers cannot treat “stability versus transformation” as an either-or decision, because the plan demands both reliable operations and rapid change at the same time.

Turning the White House AI Action Plan into operations The commentary identified agency leadership and the Office of Management and Budget as central to converting the AI Action Plan into executable programs. It argued that implementation will hinge on who holds decision rights, how accountability is assigned, and whether cross-cutting functions exist to support multiple AI efforts across an agency.

According to the article, governance, workforce development, procurement processes, and data infrastructure are the practical foundations that must be redesigned. Without these, the author warned, agencies risk repeating a pattern where promising pilots fail to reach production because no permanent capacity exists to manage AI systems over their lifecycle.

Organizing for AI at scale inside agencies

To deliver AI reliably across large agencies, the piece recommended creating standing organizational units rather than depending on ad hoc project teams. These units would carry ongoing responsibilities for AI systems, including lifecycle management and the ability to sustain programs beyond an initial demonstration.

The author also emphasized clearer governance structures for overseeing risk and performance, alongside training pipelines meant to build and retain AI talent within the federal workforce. The, as described, is to reduce fragility in delivery by embedding AI work into durable operating models rather than treating it as an exception.

Enterprise governance, procurement, and data as levers Beyond individual agencies, the commentary discussed the need to align agency governance with cross-government requirements originating from the White House and OMB. It suggested coordinated approaches for procurement, data sharing, and evaluation so agencies can meet shared standards while still adapting to mission-specific needs.

The author presented procurement reform and data infrastructure as decisive levers for scaling AI responsibly. Throughout the piece, organizational change was linked back to the AI Action Plan’s broader expectations: agencies must deliver mission outcomes while managing risk, and any structural reforms should treat the tension between stability and experimentation as a design constraint.

Stakeholders and signals to watch

The commentary said structural shifts would have immediate effects on two groups: the federal workforce and government technology vendors. It also argued that Washington-area contractors, advisory firms, and civic tech partners should expect demand to shift from short-term pilots toward longer-duration programs requiring integration and sustained support.

As next indicators, the author pointed to agency implementation plans, OMB guidance, and early organizational moves by major departments. The piece said observers should watch for concrete changes such as new staffing models, procurement authorities, and data governance commitments as evidence that strategy is being converted into structure.