The U.S. General Services Administration (GSA) plans to expand its USAi artificial intelligence evaluation platform to 16 additional federal agencies by the end of 2026, according to GSA Deputy Administrator Michael Lynch, Meritalk reported.

While speaking at the Government Service Delivery conference in Washington, D.C., on June 11, Lynch said more than 25 federal agencies are already using the platform. Launched last year, USAi provides a shared environment where agencies can test and evaluate generative AI tools without having to build and maintain separate infrastructures.

The platform allows agencies to assess AI models and applications in a secure, GSA-managed environment, conduct security and risk reviews, and evaluate performance before deploying technologies more broadly. Lynch said agencies increasingly need a safe space to experiment with AI technologies before moving them into production.

USAi was designed to help agencies scale AI adoption while reducing duplication of resources across government. Lynch noted that many technology initiatives struggle to advance beyond pilot stages, often stalling after initial testing.

Although agencies currently access USAi at no cost, GSA plans to transition the platform to a cost-recovery model beginning in fiscal year 2027, according to the agency’s budget documents.