Scout AI Raises $100M to Build AI Agents for War

Scout AI Targets Military Robotics With Reported $100 Million Raise

Scout AI, the Sunnyvale, California-based defense technology startup building what it calls the "AGI brain for defense robots," is reported to have raised $100 million in new funding, according to a TechCrunch article published April 29, 2026. The reported raise would mark a dramatic step up for a company that emerged from stealth just over a year ago with a $15 million seed round — and that has since secured multiple Department of Defense contracts and conducted a live demonstration of AI-directed lethal force. It is worth noting, however, that as of this writing, the $100 million figure has not been independently confirmed by third-party financial databases including PitchBook, Tracxn, or Crunchbase, which continue to list only the original $15 million seed round from April 2025.

Founded in August 2024 by CEO Colby Adcock and CTO Collin Otis, Scout AI has moved at an unusual pace for a defense startup — going from founding to DoD contracts to live weapons demonstrations in under two years. The company's flagship product, Fury, is a Vision-Language-Action (VLA) foundation model designed to give unmanned ground, air, and other robotic systems the ability to perceive their environment, interpret natural language commands, and act autonomously — including in GPS- and communications-denied conditions.

From Stealth to the Battlefield: Scout AI's Rapid Ascent

Scout AI came out of stealth in April 2025 with a $15 million seed round that was described as oversubscribed. The round was led by Align Ventures and Booz Allen Ventures, with participation from 13 additional investors including Draper Associates, Decisive Point Ventures, and Perot Jain. At the time of that announcement, the company had already been selected for multiple Department of Defense contracts and had built out a 20,000-square-foot research and development facility in Sunnyvale, alongside hundreds of acres of real-world testing terrain in California's Santa Cruz Mountains.

Booz Allen Ventures, the $100 million corporate venture capital arm of Booz Allen Hamilton, made Scout AI its 14th portfolio investment. The strategic nature of that relationship has since proved significant: Booz Allen is one of the U.S. government's largest defense and intelligence contractors, and its backing gave Scout AI both capital and a direct pathway into federal procurement channels.

By August 2025, the U.S. Army had selected Scout AI as one of three companies — alongside Forterra and Overland AI — to receive a share of approximately $15.5 million in contracts under its Unmanned Systems (UxS) Autonomy program. The program's goal is to retrofit existing Infantry Squad Vehicles (ISVs) with autonomous mobility capabilities. The contract spans a 16-month performance period, after which the Army may procure additional units at a total contract value of up to $150 million. Prototypes were required to be delivered to soldiers for demonstration by May 2026.

The Fury Orchestrator: AI That Directs Lethal Force

In February 2026, Scout AI conducted a demonstration at an undisclosed military base in central California that drew widespread attention — and scrutiny. During the demonstration, the company's Fury Orchestrator system directed a self-driving off-road vehicle and two armed drones to autonomously locate and destroy a target truck using an explosive charge. The demonstration illustrated, in concrete terms, what Scout AI means when it talks about giving individual warfighters command over fleets of autonomous systems.

The technical architecture behind the Fury Orchestrator is substantial. The system uses a large AI model with over 100 billion parameters for high-level command interpretation. That model — which can run on either a secure cloud platform or an air-gapped, on-site computer — issues orders to smaller 10-billion-parameter models running directly on individual ground vehicles and drones. This hierarchical approach allows a single operator to coordinate multiple autonomous platforms simultaneously without needing to manually control each one.

Scout AI's first two hardware prototypes are the G01 unmanned ground vehicle and the A01 unmanned aerial vehicle, both of which operate autonomously at the company's Santa Cruz Mountains proving grounds, powered by Fury. At the time of the February 2026 demonstration, Scout AI had four active contracts with the Department of Defense and was pursuing additional contracts, including one focused on controlling a swarm of unmanned aerial vehicles.

Adcock has described the company's ambition in direct terms. "We need to bring next-generation AI to the military. We take a hyperscaler foundation model and we train it to go from being a generalized chatbot or agentic assistant to being a warfighter," he told DNYUZ in February 2026. The company's mission, as Adcock has framed it repeatedly, is to enable what he calls "robotic mass" — large numbers of autonomous platforms operating under the coordination of a small number of human warfighters.

Why This Matters: The Race for Military AI Dominance

Scout AI's trajectory sits at the intersection of several major trends in defense technology: the rapid maturation of large AI foundation models, the push toward autonomous unmanned systems, and the U.S. military's stated urgency around maintaining technological superiority. The company is competing in a space that has attracted significant capital and government attention, with the DoD's Replicator Initiative — aimed at fielding thousands of small autonomous systems — providing a clear policy tailwind for companies like Scout AI.

What distinguishes Scout AI's approach, at least in its own framing, is the emphasis on a single unified AI model that can operate across different types of platforms — ground vehicles, aerial drones, and potentially sea and space systems — rather than building bespoke software for each. CTO Collin Otis, who previously served as a founding engineer and Director of Autonomy and AI at Kodiak Robotics and Head of Data Science and Chief of Staff at Uber ATG, has described the design philosophy this way: "Achieving warfighter-level versatility in robotic systems requires grounding AI in physical reality. By training our system on human-level behavior we make our AI embodied."

The company's growth has been notable even by the standards of the current defense tech boom. Scout AI had approximately 28 employees as of 2026, according to PitchBook data — a small team relative to the scope of what it is attempting to build. Whether the reported $100 million raise, if confirmed, would accelerate hiring and hardware production remains to be seen.

The broader context is also ethically significant. Scout AI's February 2026 demonstration — in which an AI system autonomously directed lethal action without moment-to-moment human control of each decision — raises questions that the defense and policy communities have not yet fully resolved. The company has not publicly addressed the specific question of human authorization requirements in its deployed systems, and no regulatory framework governing AI-directed lethal force in the United States has been finalized as of this writing.

Expert Reactions and Company Statements

Scout AI's leadership has been consistent and vocal about the company's strategic vision. "Physical AI is the most decisive military advantage of the century," Adcock said at the time of the company's seed round announcement. He has also articulated the commercial opportunity in the defense robotics space: "There's a very big white space for somebody to be the AGI brain for defense robots."

The company's stated goal is to reduce the human-to-robot ratio in military operations. "Our vision is one warfighter commanding many robots — seamlessly integrated into a unified team," Adcock said in a statement to Washington Technology. On the hardware and model side, Scout AI has positioned Fury as following proven design principles: "Fury is modeled on the same end-to-end learning approach that powers the most widely deployed and cost-effective autonomy systems in the world," Adcock noted in the company's Army UxS contract announcement.

From the investor side, Booz Allen Hamilton's Randy Yamada, Vice President at the firm, described the strategic rationale for the investment: "As a strategic investor, Booz Allen is hyper-focused on fast-tracking results through private sector innovation and bringing capabilities that will disrupt the federal technology market."

What's Next for Scout AI

The immediate near-term milestone for Scout AI is the delivery of autonomous Infantry Squad Vehicle prototypes to soldiers for demonstration, which was required under the Army's UxS Autonomy contract by May 2026. How those demonstrations are received will be a significant data point for both the company's DoD relationships and its broader commercial trajectory.

If the reported $100 million fundraise is confirmed through official channels, it would represent a major inflection point for Scout AI — providing capital to scale hardware production, expand its team beyond its current approximately 28-person headcount, and potentially pursue larger DoD contracts. The Army's UxS Autonomy program alone carries a potential procurement value of up to $150 million following the initial contract period, making it a meaningful long-term revenue opportunity.

Scout AI has also indicated it is pursuing a contract related to controlling swarms of unmanned aerial vehicles — a capability that would expand the operational scope of the Fury Orchestrator beyond its currently demonstrated ground-air coordination role. Whether that contract has been awarded, and under what terms, has not been publicly confirmed.

What is clear is that Scout AI has established itself, in a short period of time, as one of the more closely watched companies in the defense AI space — attracting both significant government contracts and the kind of public attention that comes with demonstrating AI systems capable of autonomous lethal action.

For more tech news, visit our news section.