A commercial trucking autonomy company just announced it is building a robot that shoots microwaves at drones. Kodiak AI and General Dynamics Land Systems revealed the Leonidas Autonomous Ground Vehicle on May 7, a fully autonomous, mobile counter-unmanned aerial system (C-UAS) platform built on a commercial Ford F-600 truck equipped with Epirus' Leonidas high-power microwave system and powered by Kodiak's own autonomous driving software. The collaboration is not a one-off prototype play, it is a formal strategic partnership positioned to bid on U.S. Army autonomous vehicle proposals. Kodiak simultaneously closed a $100 million private placement from existing and new institutional investors, including an Ares Management affiliate, signaling capital confidence in the defense trajectory.
The real story here is not the counter-UAS mission or the partnership itself. It is that a civilian autonomy company with no prior defense work is now inside a major prime contractor's bid chain, and the Pentagon explicitly enabled it. The Pentagon's published commercial-first strategy prioritizes adapting civilian technology to reduce acquisition timelines and costs, and Kodiak + GDLS is the first serious execution of that principle at scale in autonomous ground vehicles. Until now, the Army's autonomous vehicle programs have been either pure defense skunk works (Ghost Robotics, Boston Dynamics Spot variants) or prototypes that never crossed into formal procurement. Leonidas is different: it is built for a named procurement window using a commercial truck platform and a commercial autonomy stack, deliberately designed to fit the Pentagon's acquisition constraints rather than forcing the acquisition to fit the system.
The technical core is straightforward. Kodiak's driver (the software that runs the autonomous operations) handles navigation, obstacle detection, and mission execution. Epirus' Leonidas microwave system provides the counter-UAS effect, a high-power directional emitter that can disable or damage commercial drones and some more sophisticated systems without kinetic effects. GDLS serves as the lead system integrator, handling the truck integration, testing, and compliance work. The platform can operate fully autonomous or under teleoperation, and it is designed for mobile deployment across both fixed-site and perimeter defense missions. Kodiak's actual competitive advantage shows in the numbers: the company has logged more than 23,500 hours of paid driverless trucking operations as of May 7, a 120 percent increase from the end of Q4 2025. That is not a pilot metric, that is operational track record. For comparison, most defense autonomy companies count hours in the low hundreds.
What created the conditions for this announcement is a three-year convergence. First, the commercial autonomous vehicle market matured enough to prove repeatable, auditable safety and operational standards. Second, the Pentagon formalized its commercial-first purchasing doctrine in public statements and now in procurement action, after years of legacy primes failing to adapt fast enough to emerging threats like drone swarms. Third, and critically, the FY2026 NDAA increased funding for counter-UAS capabilities and explicitly authorized rapid prototyping pathways (OTAs and other transaction authority agreements) that compressed evaluation timelines from 18 months to 6. Kodiak already had the autonomy engine; Epirus had the counter-UAS tech; GDLS had the defense integrator credibility and Army relationships. The opportunity to bundle them into a formal bid did not exist two years ago.
Kodiak AI is the primary beneficiary here. The company moves from a civilian transportation market where margins are compressed and the path to profit is still uncertain, into a defense market with higher per-unit cost and smaller production runs, exactly the opposite cost structure. General Dynamics gets a faster entry into autonomous ground vehicles without building autonomy from scratch, and they preserve customer relationships with the Army at a moment when competitors like Northrop and Lockheed are also bidding. Epirus gets validation of its microwave counter-UAS approach in a formal system architecture. The Army gets a faster procurement timeline and a system built to adapt to commercial supply chains. The losers are less clear, but they exist: pure defense autonomy startups that were betting on custom platforms and extended development cycles now face a competitor with 23,500 hours of operational proof and a major prime's backing. Ghost Robotics and Boston Dynamics' defense division just got faster at decision-making if they want to stay relevant.
Three things to track: First, watch the Army's evaluation timeline for the proposals Kodiak and GDLS submit. If they move to field testing within six months of the announcement, the commercial-first model is validated. If it stretches beyond 12 months, the old acquisition inertia is still dominant. Second, monitor Kodiak's autonomous operations hours through their next earnings call, that number is the real value signal to investors and to the Army alike. If it stays flat or grows slowly, the commercial trucking business is not the engine for the defense play. If it continues doubling, the company has actual leverage. Third, watch whether other commercial autonomy companies (Aurora, Waymo, TuSimple) announce defense partnerships in the next 18 months. If they do, commercial-first is real policy. If they do not, Kodiak may have first-mover advantage that competitors cannot replicate.