Lockheed Martin’s Skunk Works bets big on AI to modernize defense projects
From AI-flown intercepts on the X-62 to 55 large language models and an OpenAI partnership, Skunk Works is rebuilding how it works
Lockheed Martin’s Skunk Works is putting artificial intelligence into the cockpit. It is also putting AI into the lab, the wind tunnel and the engineering workflow.
The X-62 flies, and the AI calls the shots
The headline result came on August 4, 2026, at Edwards Air Force Base. The X-62 VISTA aircraft carried out 27 AI-directed intercepts of a live T-38 target across eight separate flights.
The AI ran on real-time data from the Legion Pod, an infrared search-and-track sensor. The pod finds targets by their heat signatures and hands that information to the AI, which decides what the aircraft should do next.
That is what engineers mean by closing the “sensor-to-action loop.” The machine sees something, interprets it and acts on it, all on an operational combat platform rather than in a simulator.
Ron Fehlen described the X-62 tests as a decisive step toward AI-augmented air dominance for the United States.
AI, tech, and the markets they move—in one daily briefing.
Daily. Free. Join 34,000+ readers across crypto, finance, and policy.
Three months, one framework called Supermassive
Skunk Works integrated and ground-tested the AI agent for the X-62 in just three months. The division credits a development framework it calls “Supermassive.”
55 language models and an OpenAI partnership
As of October 2026, Skunk Works was using 55 distinct large language models. Lockheed has also partnered with OpenAI, with the collaboration targeting advanced mathematics and physics challenges connected to improving the F-35’s sensor technologies.
Vectis gets a faster runway
On September 14, 2026, Lockheed announced plans to build four Vectis prototypes, with the first flight expected by the end of 2027. Wind-tunnel testing wrapped in under two months.
Why the Pentagon is paying attention
The Defense Department wants stronger collaborative combat aircraft and better human-machine teaming. Collaborative combat aircraft are uncrewed systems meant to operate alongside crewed jets.
The X-62 intercepts speak directly to that vision. An AI that can track and pursue a target using live sensor data is a building block for drones that fly in formation with human pilots.