Army Funds Three Autonomy Startups to Field Combat Vehicles by May

A three-officer squad from Fort Moore, Georgia, will load their equipment onto an Infantry Support Vehicle in May 2026 and watch it navigate rugged terrain without a driver. That vehicle will be running autonomy software from one of three commercial startups that the U.S. Army just funded to prove their technologies work in actual combat conditions. The Army awarded approximately $15.5 million in OTA agreements—about $5 million per company—to Forterra, Overland AI, and Scout AI to integrate their commercial autonomy solutions onto Army ISVs and deliver prototypes for demonstration and evaluation alongside soldiers from a Transforming in Contact (TiC) unit during Combat Training Center rotations. This is not a lab test. This is not a contractor briefing to Army leadership at a conference. Soldiers will operate these vehicles. The Army has signaled it will make acquisition decisions based on what those soldiers report.

The context matters here because it explains why this OTA structure is different from the usual prototype funding cycle. The ISV program itself is not new—the Army approved full-rate production three years ago with an acquisition objective of 2,600 vehicles. As of November 2025, more than 1,100 had been delivered to mobile brigade combat teams, airborne units, and a Ranger regiment. That leaves roughly 1,500 more vehicles to be manufactured over the next five to seven years. The three autonomy OTAs are structured specifically to determine whether one of these commercial startups can own that remaining production wave. A successful May demonstration could trigger a follow-on production OTA that scales from $5 million in prototype funding to something closer to $200 to $400 million in production contracts, depending on how aggressively the Army commits to autonomous ISVs across all formations. The structure is brutal in its clarity: demonstrate or lose access to the entire production pipeline.

Each awardee is taking a different technical approach. Overland AI, based in Seattle, is emphasizing off-road maneuver performance and autonomous navigation through the kind of terrain a mechanized infantry unit would encounter during combat training center exercises. Scout AI, in Sunnyvale, is delivering its Fury software stack integrated with Textron Systems hardware and validated by Edge Case Research—a third-party safety validation firm that has become essential to military autonomy contracts because it reduces the Army's liability exposure on untested systems. Forterra, based in Clarksburg, Maryland, is deploying its AutoDrive self-driving technology directly onto the ISV, with the company stating the system will enable vehicles to 'operate independently or as part of a connected force,' ensuring 'seamless interoperability, resilient communications, and robust autonomy' across Army formations. The Army is not picking a single architectural approach and funding three companies to build it faster. It is funding three completely different autonomy stacks and will use soldier feedback in May to decide which one actually works in the field. That is a significant distinction because it means the Army is willing to tolerate technical diversity in exchange for speed and learning—a structural change from how military contracts usually work.

Why compress the timeline this aggressively? The framing came directly from Emil Michael, the Under Secretary of War for Research and Engineering, who stated flatly: 'What we've learned from the Ukraine-Russia war is that the front lines of a conflict over territory are robot-on-robot now.' That is not aspirational language. That is operational doctrine. The Pentagon's Drone Dominance Program is already allocated $1.1 billion for unmanned systems across all domains, and Owen West, confirmed as DIU Director in March 2026, has made clear that the new priority is speed, scale, and lethality—meaning companies that can demonstrate hardware in the field within months, not years, will capture the investment flow. The ISV autonomy OTAs are structured as a direct test of that principle: can three commercial startups produce soldier-ready autonomous vehicles in six months? If not, the Army will narrow its focus and potentially award the entire production pipeline to a single vendor that can scale faster. If yes, the May 2026 demonstration becomes the gating event for production volume allocation.

Who benefits from this structure, and who doesn't? The clear winner is whichever startup delivers the most operationally convincing autonomous ISV in May—the company that soldiers trust enough to ride alongside uncrewed vehicles in a contested terrain exercise. That company gets first access to the remaining 1,500 ISV production slots, which translates to a multi-hundred-million-dollar production contract over five to seven years. Overland AI appears to have the strongest technical positioning for this specific mission because it has been building off-road autonomous systems specifically for rough-terrain operations—precisely the conditions that the Army's TiC units will be testing in May. Scout AI has the safety validation advantage through Edge Case Research, which matters significantly if the Army is risk-averse about deploying uncrewed vehicles with soldiers nearby. Forterra is less known but is betting that integrated self-driving technology can outperform point solutions. The real loser in this scenario is any company not selected for the May demonstration—because the demonstrated results will define the technical baseline that the Army uses to evaluate follow-on production awards. A startup that fails to make the OTA cut will find it extremely difficult to catch up when the production contract is awarded, because the Army will have already trained soldiers on a specific autonomy stack and will not want to revalidate a new system.

Here is what is actually happening beneath the announcement: the Pentagon is using OTA agreements to collapse the acquisition timeline and force commercial defense startups to bet their entire company on a six-month engineering sprint. This is not new in principle—OTAs have been used for exactly this purpose since the DIU was created—but the scale and the explicitness of the ISV case is new. The Army is not saying 'develop a prototype and we will evaluate it sometime in the future.' It is saying 'you have six months to integrate your autonomy stack onto the ISV, deliver functioning vehicles, and demonstrate them to soldiers in a live training exercise that will directly inform production acquisition decisions.' That is a capital and engineering intensity that only well-funded startups can sustain. It screens out undercapitalized companies and companies without proven hardware platforms. And it compresses risk: instead of spending three years and $50 million on a program that might never transition to production, the Army is forcing a 180-day prove-it cycle that makes follow-on acquisition decisions obvious. The startup that loses in May does not get a consolation prize. It gets a $5 million contract that expires and no path to scale.

The timing also matters. The May 2026 demonstration is scheduled for a TiC unit rotation at a Combat Training Center—meaning real soldiers, real terrain, and real operational scenarios. The Army is explicitly saying it will not evaluate autonomy in a controlled lab environment or against a pre-written test matrix. It will evaluate autonomy by watching soldiers operate the vehicles during actual training exercises and asking those soldiers whether the system is trustworthy, whether it performs as promised, and whether it enhances squad mobility without creating safety vulnerabilities. That is a much higher bar than a government test and evaluation event, and it is designed to eliminate false positives. A company can game a spec sheet. You cannot game soldier feedback from a TiC rotation. The fact that the Army structured the OTA this way—demonstrating directly to soldiers rather than to Army evaluators—signals that the December 2025 appointment of Owen West as DIU Director has already changed how the Pentagon evaluates military technology. West has been explicit about deprioritizing contractor briefings and long acquisition cycles in favor of companies that can ship hardware fast and prove it works in the field. The ISV autonomy OTAs are the institutional instantiation of that policy shift.