Scout AI raises $100 million to build military AI models

On Wednesday, defense startup Scout AI announced its latest funding and contract milestones as it builds software to control autonomous military vehicles. Founded by Colby Adcock and Collin Otis, the company is positioning itself to develop AI models specifically for conflict zones.

The company’s financial and contract details include:

• Series A Round: $100 million, led by Align Ventures and Draper Associates, announced on Wednesday.
• Seed Round: $15 million, raised in January 2025.
• Military Contracts: $11 million in technology development contracts secured from organizations including DARPA (the Defense Advanced Research Projects Agency) and the Army Applications Laboratory.

Scout AI is developing an AI model called “Fury” to operate and command military assets. The company is utilizing Vision Language Action (VLA) models—AI models used to control robots—which build on existing Large Language Models (LLMs). Unlike autonomous cars that operate in structured environments, VLAs allow models to learn from physical interactions. Scout AI has been training its models using all-terrain vehicles (ATVs) for six weeks at a military base in California, where an ATV carried passengers on a 6.5 km loop under autonomous control.

CTO Collin Otis explained the value of VLAs: “You’re actually just learning how to connect your prior knowledge to these couple little joysticks. It’s not a big leap. That’s the way to think about VLAs and why they’re such an unlock.”

Co-founder Colby Adcock expects the startup’s first widely adopted product to be “Ox,” a command and control software. In the years ahead, Scout AI plans to build its own model from the ground up. Otis noted that interacting with the physical world is key to developing Artificial General Intelligence (AGI), arguing that reading the internet only provides limited intelligence.

The startup is currently working with the U.S. Army’s 1st Cavalry Division, which is using technology from 20 autonomy companies during its training cycle at Fort Hood in Texas. The unit is next expected to deploy in 2027, with the expectation that it will bring along products that prove themselves.

Scout AI is focusing on the defense sector at a time when many commercial AI developers avoid military partnerships. Otis noted that many AI developers are reluctant to work with the military, particularly when it comes to running agents on weapons systems. While autonomous weapons remain politically sensitive, Scout AI is also practicing with drones for reconnaissance and defense, testing systems where munition drones could search specific geographic areas.

Why it matters

Scout AI is positioning itself as a “frontier lab for defense” by building an intelligence layer for military machines. By using Vision Language Action models, the company aims to train AI systems to operate in unpredictable conflict zones, marking a shift toward autonomous military assets.