How might we enable autonomous and unmanned systems so that the U.S. can project defense readiness with lower cost, faster deployment, and greater adaptability?

U.S. military superiority has traditionally come from platform quality: better aircraft, better ships, better sensors. But this approach is increasingly challenged by adversaries who can field large numbers of lower-cost systems. China is building ships faster than the U.S. can. Drones costing thousands of dollars can threaten aircraft costing tens of millions. Swarms of autonomous systems can overwhelm defenses designed for smaller numbers of high-value targets. The U.S. can't match adversary production capacity in a quantity race — but it can leverage advantages in AI, autonomy, and system integration to field unmanned systems that are more capable, more adaptable, and more cost-effective than adversaries' piloted platforms.

This requires a fundamental shift in how the U.S. thinks about military capability: from exquisite platforms operated by highly trained personnel to networked systems where autonomy multiplies human decision-making. Unmanned aircraft that can operate in contested airspace without risking pilots. Autonomous surface vessels that can patrol vast ocean areas persistently. Ground robots that can clear routes or conduct reconnaissance. And AI systems that can coordinate hundreds of assets simultaneously, adapting to adversary actions faster than human commanders could direct individually. The challenge is making this work within legal and ethical constraints, maintaining human oversight while enabling speed-of-machine decision-making, and building systems that adversaries can't easily counter or subvert.

Unmanned systems entered military use gradually over several decades. The first military drones were developed in the 1960s for reconnaissance, providing real-time intelligence without risking pilots over hostile territory. Early systems like the Ryan Firebee were essentially remote-controlled aircraft — a human pilot controlled them from the ground, just not from inside the aircraft. This approach worked for simple missions but couldn't scale because each aircraft required a dedicated operator and communications link.

The breakthrough came from increasing autonomy. The Predator drone, introduced in the 1990s, could fly waypoint missions autonomously, requiring human intervention only for takeoff, landing, and targeting decisions. This meant one crew could operate multiple aircraft sequentially, improving efficiency. The addition of weapons capability in the early 2000s made Predator a strike platform, not just an intelligence collector. By the peak of operations in Iraq and Afghanistan, the U.S. was flying dozens of armed drone missions daily, demonstrating that unmanned systems could perform complex missions traditionally requiring piloted aircraft.

But these systems still depended on continuous communications links and human oversight for critical decisions. This worked in permissive environments like counterterrorism operations, where adversaries couldn't jam communications or contest airspace. It wouldn't work against peer adversaries like China or Russia, who could jam communications and threaten drones with air defenses. This created demand for more sophisticated autonomy: systems that could operate in denied communications environments, adapt to adversary actions independently, and coordinate with other systems without continuous human direction.

AI advances in the 2010s-2020s made higher levels of autonomy feasible. Machine learning enabled drones to navigate without GPS, using visual odometry and terrain recognition. Computer vision allowed automatic target recognition, distinguishing vehicles from buildings or civilians from combatants. Multi-agent reinforcement learning enabled swarms of drones to coordinate autonomously, adapting to changing situations without human micromanagement. And AI planning systems could generate mission plans, optimizing for objectives while respecting rules of engagement and operational constraints.

Ukraine's war with Russia demonstrated both the potential and the challenges of autonomous systems. Ukrainian forces used commercial drones extensively, often modified with AI for navigation and targeting. Drone strikes on Russian ships, aircraft, and logistics demonstrated that low-cost autonomous systems could achieve strategic effects. But Russia adapted by deploying jamming, shooting down drones, and developing their own autonomous systems. This sparked a rapid co-evolution: each side developed new autonomous capabilities, and each side developed countermeasures. The lesson: autonomy is not a static advantage but a domain of continuous competition where the advantage goes to whoever can innovate and adapt fastest.