Colby Adcock’s Scout AI raises $100M to train its models for war. We visited its bootcamp
At a secret military base, Scout AI's ATVs learn to navigate war zones autonomously.
Scout AI, a defense startup founded in 2024 by Colby Adcock and Collin Otis, has raised $100 million in a Series A round led by Align Ventures and Draper Associates, following a $15 million seed round in January 2025. The company is building an AI model called Fury, based on Vision Language Action models (VLAs) from Google DeepMind, to operate and command military assets in unpredictable conflict zones. Scout has secured $11 million in contracts from DARPA, the Army Applications Laboratory, and other DoD customers, and is one of 20 autonomy companies testing with the U.S. Army's 1st Cavalry Division at Fort Hood, Texas.
Scout's training involves autonomous ATVs navigating challenging off-road terrain at an undisclosed military base in central California. CTO Collin Otis, formerly of autonomous trucking firm Kodiak, compares training Fury to teaching soldiers, starting with base intelligence from existing LLMs and refining it for military-specific tasks. Colby Adcock, who also sits on the board of humanoid robot company Figure.AI (led by his brother Brett), says the VLA approach allows vehicles to adapt quickly to new environments, unlike traditional autonomous systems that rely on structured rules. The company expects its technology to prove itself by 2027, when the 1st Cavalry Division next deploys.
- Scout AI raised $100M in Series A funding, led by Align Ventures and Draper Associates, for its Fury AI model
- Fury uses Vision Language Action models (VLAs) to train autonomous vehicles for off-road combat logistics and weapons
- Scout has $11M in DoD contracts and is testing with the U.S. Army's 1st Cavalry Division at Fort Hood
Why It Matters
Scout AI's Fury model could redefine military logistics and autonomous warfare, raising critical ethical questions about AI in combat.