How Meta Is Supporting US National Security With AI

Since making Llama AI models available to US government agencies and their contractors, a wide variety of organizations have adopted this technology to bolster our nation's security. Today, dozens of industry stakeholders use Llama to empower America's military and intelligence community with advanced, bespoke, and secure tools for their critical work.

Partners like Amazon Web Services and Snowflake are expanding access to Llama for America's military branches in highly sensitive environments and networks. Other partners are leveraging Llama's capabilities to address the needs of service members and to develop cutting-edge products that advance their missions.

Accelerating Innovation

Organizations can use Llama to securely build and deploy new applications tailored to their highly specialized needs, while retaining complete control over both the model and their data within their own infrastructure thanks to its open source release. This enhanced security and adaptability is key in the national security space and makes Llama an ideal solution for startups looking to build novel tools in a resource-efficient way.

Powering Special Operations

Defense startup Legion Intelligence collaborated with US Special Operations Command (USSOCOM) to launch SOFChat: USSOCOM's first enterprise-wide generative AI platform, bringing Llama to American special operations forces. Legion's integration of Llama into SOFChat is changing how service members work by providing machine-speed analysis and real-time monitoring of conditions on the ground. With SOFChat, special operations forces are able to generate intelligence reports 18 times faster and process video footage nine times quicker, giving them the speed and insight they need to make better and faster decisions when every second counts.

Deployed at the Edge

Llama can be fine-tuned with secure data and deployed directly onto local devices, enabling reliable use anywhere, including remote and extreme environments. This flexibility makes Llama an ideal solution for supporting American personnel deployed in the field, who often operate in challenging conditions where internet connectivity is limited or unavailable.

Supporting Military Missions

EdgeRunner AI has developed a Llama-based language model that can run on consumer-grade laptops, allowing service members to access cutting-edge AI even in adversarial environments when being disconnected becomes mission critical. From identifying safe locations to land an aircraft or accurately translating languages, to quickly calculating food and water needs, Llama is supporting military missions without sacrificing data privacy or operational security.

Driving Efficiency

Lockheed Martin is using Llama to support training and flight simulation efforts for the US Air Force, Navy, and international pilots. By using Llama within Lockheed Martin's AI Factory platform, the Skunk Works® team created a virtual subject-matter expert that enabled software engineers to quickly understand the principles of the government's Joint Simulation Environment (JSE) and develop compatible code to rapidly integrate models. Llama was crucial in aiding engineers to develop capabilities more rapidly in areas where they had limited prior expertise.

These are just a few examples of the many ways Llama supports American national security. We look forward to continuing to support the US government and its allies as they use AI to enable discoveries and breakthroughs, drive growth and efficiency, and help keep our nation safe.