In a September exercise on a Florida airfield, members of the 10th Special Forces Group launched a drone assault that mirrored the “spiderweb” attack that Ukraine had recently staged against Russia. The defenders were counter-drone troops from across the U.S. military, trained for a week on tech that the Pentagon has spent billions to develop. 

U.S. counter-drone efforts haven’t been the same since.

“What I would tell you is that it helped us develop our priorities,” said Brig. Gen. Matt Ross, who leads Joint Interagency Task Force 401, the Pentagon’s counter-drone clearinghouse.

Dubbed Operation Clear Horizon, the exercise at Eglin Air Force Base sought to replicate conditions and weapons seen on the battlefields of Ukraine.

The special operators “had lessons learned from Ukraine in Eastern Europe and they came back and they said, ‘This is what we're seeing on the battlefield’,” Ross said.

In their mock assault, the special operators used a wide range of drones, from small to large, and many that were resistant to jamming and radar.

“They flew drones that were regular [radio-frequency], commercial drones. They flew drones that had directional antennas on them so they're harder to jam, drones that were frequency hopping so they have a more resilient connection” against electromagnetic attack, Ross said Tuesday at the Sea-Air-Space event. “They went up to Group 3 drones” and down to Group 1s. “We used fiber optically-controlled drones…We used drones controlled by LTE, the cellular network,” enabling operators in Colorado to launch against targets in Florida, a first for the U.S. military.

That effort to replicate the battlefield of Ukraine is a big departure from the way the department usually tests its technology against drones.  

Because the electro-magnetic effects used against drones interfere with airplane guidance and cellular service, the military can only test them in very particular circumstances. Even the most recent counter-drone tests—such as August’s T-REX exercise at Camp Atterbury, Indiana, and the Army’s FlyTrap exercise last November in Germany—don’t do so. Instead, they mostly test concepts for bringing down drones without using million-dollar missiles. 

What’s more, data from such tests don’t usually fully inform the planning of other exercises and experiments, according to participants and observers of such events. Between September and December, Ross said, 67 tests were conducted by the services, combatant commands, the Pentagon’s research office, and other DOD outfits. 

“It was all well-intentioned. But we couldn't see all of that data in a way that would allow informed comparisons between different systems that were tested at different venues,” he said.

Even if circumstances make it hard to test jamming defenses against UAVs, the military still needs to drill against drones built to elude radar as well. What the U.S. really needs now is continuously up-to-date understanding of how U.S. tactics and gear would perform on a battlefield like Ukraine. That’s what the task force is trying to do.

“We were over in Ukraine about six weeks ago, talking to the Unmanned Systems Force [sic], watching how they defend Kyiv on any given night, understanding what they have along the forward line of troops,” Ross said. “Then we've looked at those most promising technologies and we referenced their performance data in Ukraine instead of internal department testing and evaluation.”

That’s leading to changes in how the U.S. buys equipment to take down drones and plans defenses for installations and forces. 

The September exercise in Florida showed that U.S. drone defenders needed a way to combine the data coming in from far-flung radars, drones, and counter-drone systems.

“If you look at the Department of War installations across the panhandle of Florida, we should be able to identify a [drone] track from the west and pass it between installations,” Ross said.

Now, the U.S. has a single drone-tracking software solution and interface across the services, he said. (Defense One has asked for details.)

“If you were to go to a location where you have multiple services working together, and even other federal agencies or international partners, we have seamless air domain awareness and the ability to sense connect any sensor with any effector”—that is, a drone-downing system, Ross said.

The exercise also revealed a need to focus more on long-range drones that can damage  “high-payoff targets, which are going to be command-and-control, logistics, or air defense,” he said.

For smaller drones—Groups 1 and 2—the U.S. must develop interceptor drones that cost less than today’s expensive defense missiles.

Ukraine has already absorbed such lessons, which the task force is now passing to U.S. commanders, including those in the Middle East. 

“We've procured some of those systems now to start integrating across the Department of War,” he said. “In the past six weeks, we've committed over $600 million to this problem, specifically for the rapid integration of new counter [unmanned aerial system] technology.”

In the 2027 budget proposal, the Pentagon is requesting $75 billion for new drone technology, a sum larger than the annual GDP of some countries and the current budget of the U.S. Marine Corps. 

If those numbers seem far apart, it’s because the United States has highly-effective missiles to take down drones. But they were designed to take down missiles, and they’re expensive. But Ross says that getting the most out of both efforts means linking them together in a way that traditional long-range fires teams and missile-defense units have not been, historically. 

Offensive and defensive drone operations are “inextricably, inextricably linked,” he said. 

The threat, however, is evolving far faster than annual budgets. Ross took issue with the idea, advanced by some senior military leaders, that today’s drones are analogous to the IEDs of the Iraq and Afghanistan wars. No commercial market spurred the evolution of IEDs, nor could software and data spur improved performance.

“For IEDs, you had no commercial application for that technology. With unmanned systems, and specifically with autonomy, there's so many commercial applications that we're going to see this accelerated development in this space,” he said. “That's going to cause us concerns, from a security perspective.”

]]>

The White House Office of Science and Technology Policy on Thursday accused China and other foreign entities of engaging in “deliberate, industrial-scale campaigns to distill U.S. frontier AI systems,” and said that the Trump administration will be taking steps to safeguard domestic artificial intelligence products. 

In a memo to federal agencies, the White House office warned that these distillation campaigns — in which an attacker sends a deluge of requests to an AI model to train a knockoff version of it — are allowing bad actors to steal proprietary information from U.S. companies.

“Models developed from surreptitious, unauthorized distillation campaigns like this do not replicate the full performance of the original,” the memo said. “They do, however, enable foreign actors to release products that appear to perform comparably on select benchmarks at a fraction of the cost.”

Anthropic in February accused three Chinese-based AI companies — DeepSeek, Moonshot AI and MiniMax — of overwhelming its Claude model with 16 million exchanges from roughly 24,000 fraudulent accounts. 

Those allegations came the same month that OpenAI sent a letter to members of the House China Select Committee that said, in part, that it had seen evidence “indicative of ongoing attempts by DeepSeek to distill frontier models of OpenAI and other US frontier labs, including through new, obfuscated methods.”

Thursday’s memo does not cite any specific companies engaged in distillation campaigns against U.S. AI firms. But OSTP Director Michael Kratsios said in an X post that “these foreign entities are using tens of thousands of proxies and jailbreaking techniques in coordinated campaigns to systematically extract American breakthroughs.”

OSTP told agencies that the Trump administration will be taking a series of steps to expand engagement with U.S. companies and crack down on foreign-based distillation campaigns. 

These include sharing more information with the private sector about attempts to conduct large-scale distillation attacks, enabling companies “to better coordinate against such attacks;” partnering with firms to develop a set of best practices to counter these campaigns; and looking at developing new steps to hold foreign actors accountable for their actions. 

The memo said these actions are consistent with the White House’s AI Action Plan, which was released in July 2025 and emphasizes the importance of “preventing our adversaries from free-riding on our innovation and investment.”

The White House’s warning about China-based distillation campaigns is the latest salvo in the U.S. and China’s ongoing competition to lead the global AI race. It also comes as major American AI firms have rolled out what they say are advanced AI models that have exquisite cybersecurity capabilities that could cause national security risks if they fall into the wrong hands.

Retired Gen. Paul Nakasone, who led the NSA and U.S. Cyber Command, said the administration may consider export controls, diplomatic protests and tailored technology restrictions as potential responses to the distillation efforts.

“And we’re going to be very, very careful about how we’re going to share that [AI technology] with a series of different partners,” he said, speaking at a Wednesday roundtable with reporters in Nashville when asked about the campaigns. Nakasone now leads Vanderbilt University’s Institute of National Security.

Given China’s increasingly bellicose tone toward Taiwan, and the potential for preemptive actions against the U.S. in advance of a full-scale invasion of that country, lawmakers have also been worried about how technology advances will ultimately benefit Beijing. Through China’s military-civil fusion strategy, the country has moved to enhance its military strength by removing barriers with its commercial sector.

President Donald Trump and Chinese President Xi Jinping are scheduled to meet next month in Beijing for a summit to discuss a host of issues, including export controls on semiconductors and IP theft.

]]>