US Federal Government Reaches Landmark Partnership with Meta, Officially Adopting Open Source Llama AI Model to Serve the Public Sector

September 24, 2025
GSA
4 min

Abstract

The U.S. General Services Administration (GSA) announced on September 22 a landmark partnership agreement with Meta, integrating Meta's open-source artificial intelligence model, Llama, into the federal government's OneGov program. This initiative will provide all federal departments and agencies with more convenient access to this AI tool. This move eliminates the need for individual agencies to negotiate separately, saving time and reducing duplication of effort within the federal government.

This arrangement supports the goals of President Trump's American AI Initiative and the requirements of Office of Management and Budget (OMB) memoranda M-25-21 (Accelerating Federal AI Use Through Innovation, Governance, and Public Trust) and M-25-22 (Driving Efficient Procurement of AI Across Government).

Key Takeaways:

  • As an open-source model, Llama allows federal agencies to maintain full control over data processing and storage.
  • Since the model is publicly available, technical teams can build, deploy, and scale AI applications at a lower cost.
  • In recent months, GSA has also approved AI tools from competitors including Amazon Web Services, Microsoft, Google, Anthropic, and OpenAI.

Policy Context and Significance

Josh Gruenbaum, Federal Acquisition Service Commissioner, stated: "In alignment with President Trump's AI Action Plan, GSA is committed to integrating AI into government operations. Thanks to Meta and its commitment to advancing public sector services, our federal partners now have easier access to Llama and open-source AI."

Mark Zuckerberg, Founder and CEO of Meta, commented: "America leads in AI, and we want to ensure all Americans see the benefits of AI innovation through better, more efficient public services. Through Llama, U.S. government agencies can better serve the people."

Technical Features and Application Scenarios

Llama is a free large language model capable of processing data including text, images, video, and audio. Federal agencies will be able to deploy this tool to accelerate tasks such as contract review or resolving information technology malfunctions more quickly.

This OneGov arrangement is unique because Meta is providing its Llama model for free. GSA focuses on the backend work, validating that Llama meets federal requirements and providing consistent, streamlined access across government.

Competitive Landscape and Industry Impact

The federal government has accelerated AI adoption in recent years, driven by the potential to reduce costs and improve efficiency within its vast bureaucratic system. The Department of Defense (DoD) has been a major driver of federal AI spending, accounting for 95% of the total potential contract value in 2022.

Data shows that federal AI contract obligations surged from $261 million to $675 million in just one year, while the value of potential future awards soared by nearly 1200% to $4.56 billion.

Meta's approval comes just weeks after Amazon-backed Anthropic offered its Claude AI to U.S. agencies for a symbolic price of $1. OpenAI made a similar offer earlier this year, pricing ChatGPT Enterprise at $1 per agency per year for the first year.

Data Security and Compliance

By partnering with Meta to ensure federal agencies can access Llama according to government needs and legal requirements, this collaboration strengthens the federal government's ability to test, adapt, and deploy AI while maintaining full control over sensitive data.

Open models like Llama provide agencies with the flexibility to build task-specific solutions, reducing reliance on closed providers and promoting transparency and reproducibility in AI research.

Future Outlook

When asked whether tech executives were offering discounts to curry favor with President Trump, Gruenbaum responded: "This isn't about currying favor, it's about recognizing how we come together to make this country the best country it can possibly be."

Industry observers believe that the government's embrace of a wider range of AI providers will encourage healthy competition while reducing dependence on any single vendor. This approach also allows agencies to select the model best suited for specific tasks, rather than relying on a one-size-fits-all solution.

This partnership agreement marks a significant step for the U.S. federal government in AI adoption, not only reflecting its emphasis on open-source technology but also its strategic considerations in maintaining technological leadership while ensuring data security and cost-effectiveness.