Lars Erik Schönander is a nonresident fellow with FAI’s tech and statecraft team and an MBA student at Dartmouth’s Tuck School of Business. From 2022-23, he wrote a series of articles and reports evaluating the extent of American dependence on Chinese drones and making the case that this dependence posed serious national security risks.
In 2022, Lars penned an op-ed for The Wall Street Journal on U.S. federal agencies’ overreliance on Chinese drones. The following year, he undertook an extensive FOIA endeavor to understand state-level dependence on Chinese drones, compiling these findings in an FAI report, and wrote for American Affairs on the need for a domestic drone industrial base, among other work.
Three years later, the drone policy picture has changed tremendously. Congress has broadly restricted federal procurement of Chinese drones and the FCC has placed foreign-made drones and drone components on its Covered List of banned equipment. On the demand side, the Pentagon has announced major purchasing commitments and the administration has directed agencies to prioritize buying domestically manufactured drones. Drones have now been deemed too risky to national security to be built elsewhere.
I spoke with Lars in early October about how this shift came about and the lessons we should take from it. We discussed what the drones case can teach policymakers about eliminating dependencies and growing a domestic market in other industries, his experience as an early adopter pushing for action on Chinese supply chain risks before our government was well-versed in dealing with them, and the Chinese drone juggernaut that is DJI.
This interview has been edited for brevity and clarity.
When you began writing about the Chinese drones problem in 2022, DJI drones dominated the commercial sector and were used widely by government. Some agencies had woken up to the risks this posed and banned their use already—the Army in 2017, Interior grounded its non-emergency fleets in 2020—but there had been no congressional action broadly limiting their commercial use or federal procurement, save for DOD use.
Against that backdrop, how did you decide to work on this issue and which angle to take?
I already had a background using public records to bring attention to China issues, and decided to focus on drones next after learning about DJI. I found it very troubling that essentially one company had a global monopoly on small commercial drones.
Through record requests at the federal level I learned that, despite multiple agencies saying they stopped using Chinese drones, some of them—like the Secret Service or USDA—continued to use these drones. That turned into an op-ed for the Wall Street Journal on federal agencies using Chinese drones, which turned into a bigger project of figuring out just how many Chinese drones were being used in the United States at the state level.
Why focus on the state level?
A lot of the fight was already at the federal level, but new evidence would be at the state level because that’s where the bulk of the civilian fleets were—think forestry agencies for fighting fires, transit agencies for conducting road surveillance work, agriculture agencies for land management, and fish and wildlife agencies for monitoring animals.
It had already been established that most police agencies were using Chinese drones. I was looking at other state and local agencies that haven’t been tackled as much. I found a similar problem. I filed public-records requests with over 100 state agencies, usually their agricultural or transportation agencies. Through all those requests I found that, across agencies, an average of 85% of drones at the state level were from China, with the vast majority of those being from DJI.
What is the actual risk of a state forestry agency using Chinese drones?
The issue is not necessarily specific sensitive data in all of these cases, but that the data they gather might be sent to the Chinese government. As a Chinese company, DJI is mandated by their National Intelligence Law to share data with the Chinese government when asked. The data is often stored in servers in China, so you don’t control your own data and you don’t know where it’s going. You don’t know if, once you connect it to your own internet, someone has been fumbling around your internal network trying to find something out.
It’s a similar reason to why people are concerned about Chinese cranes. They’re connected into the internal IT and you simply don’t know if someone has the access to parts of a network and information that you don’t want them accessing.
It’s also a supply chain issue. DJI is the only game in town. That means if, for instance, there was a conflict, up until recently we would have been in a very difficult situation with drone supply. Because while the United States is good at manufacturing big drones like Predators, those are really expensive—state and local governments need smaller cheap drones that can do the job for a few thousand dollars.
When reading your articles, it struck me as concerning that some agencies started restricting the use and purchasing of Chinese drones in 2017, but Congress didn’t limit federal procurement across the board until 2023, with the American Security Drone Act.
Why do you think there was such lag between parts of government recognizing the risk and comprehensive action? Was it foreign lobbying or influence, general congressional inertia, a lack of American-made alternatives, or some combination of these factors?
There is an angle with DJI lobbying. They had a very big lobbying arm and were associated with various American companies for a while before they slowly started getting removed as clients from those lobbying firms.
Part of the challenge was that China policy was pretty slow going until after COVID-19. It takes time for these issues to go from think tank and advocacy world to Congress actually passing something like the American Security Drone Act, which was introduced in 2019 and 2021, but only passed in 2023. And even once the idea arrives, it’s hard to get Congress to shift from “this is troubling” to “I should care that this is troubling.” That’s the big challenge. I think this came to drones early because of Ukraine. People were aware of DJI, but it wasn’t until they discovered that a small drone can blow up exquisite systems worth millions of dollars apiece that they began to really care.
Then there’s the instance of companies behaving weirdly. There was an incident with random drones flying over the White House and several restricted places across DC. In 2022, hundreds of recreational DJI drones flew through restricted airspace in DC. This got members of Congress, including Marco Rubio, to ramp up investigations of Chinese drones and take the problem more seriously because they saw it at the federal level.
The main reason why these drones took over our market so quickly and why it took so long to get them out is that DJI is incredibly cheap. What allowed them to do this, and do you think American companies are positioned to become cost-competitive?
First, DJI drones were more feature friendly. They have often been called the “Apple of Drones,” and that’s a pretty reasonable assessment for how well they handle. They started in Shenzhen in 2006 with 10-20 employees, making drone parts only, and within a few years were making their own drones.
I think this is a good example of the functioning industrial commons that Shenzhen is. It’s one of the electronic capitals of the world when it comes to manufacturing. There’s so much knowledge and talent around there that they’re able to hire rapidly to get the skills they need and find the markets they need to sell their products.
They did also get Chinese funding, which they hid, and that certainly helps. But, in all fairness, there are many Chinese state-backed companies that don’t do well globally. A lot of money helps to subsidize the product, but they had to have a good product people wanted to buy in the first place.
When it comes to whether the U.S. will be able to compete on price with DJI, we are making more and more drones, so the price point will come down. I wish I could say definitively yes, but I think DJI at the price point as it is is still going to be lower than American companies, especially when it comes to entry-level and mid-range consumer drones, where we aren’t even competing.
How would 2022 Lars feel about our progress on drones? Where have policy and outcomes exceeded your expectations and where have they fallen short?
I think he would have been excited that they finally moved on legislation—the American Security Drone Act of 2023—to do the bans after years of going back and forth on it. There’s finally legislation that’s been signed by a president out on paper.
There’s also been really good progress from the Department of Defense, where over the past year or so they have stood up an initiative called the Drone Dominance Program. The Department’s been running competitions with American drone makers to see how well they perform, with the winners getting massive purchase orders from government so that companies can scale.
Where American Affairs 2023 Lars would not have been satisfied is that work still needs to be done in industries upstream of drone production. Part of the reason DJI is so successful is that the components are also cheap and made in China — for example cheap actuators and cheap motors. DJI designs and manufactures its own drone motors in-house.
We’re starting to solve this problem. In the American Dynamism portfolio, there’s a new startup called Westmag, and they’re trying to solve the motor problem for drones and the actuator problem for robots. The problem is you can’t just focus on the defense industrial base. You need to have a full regular industrial base before the defense industrial base can work well, because you need all these small, unsexy industries working properly, and you need enough talent to enable building these types of products.
Which lessons should policymakers take away from the Chinese drones saga for other sectors where we are similarly trying to decouple from China while building a domestic market—for instance LiDAR or mobile robots?
Drones are a good example for why it matters to get the regulatory environment right. Essentially, for a while it was a lot harder to fly a drone in the United States versus in China. And the problem with that is while there are some more restrictions these days in China over where drones can fly, it took a lot longer for those to come about. So we simply had fewer opportunities for people to practice flying commercial drones and using commercial drones than China.
The lesson for industrial policy broadly is learning how to work proactively versus reactively. By the time we discovered that drones were a problem in the United States, we essentially had no commercial American-made drone industry. So it’s a good reminder that one needs to be looking ahead to figure out what the next risky industries are and who’s competing in them. We want to avoid situations where in five years we suddenly identify a problem and the only response is to spend a billion dollars and ban competitors.
To that end, are there any industries that you’re eyeing now as potentially risky and in need of more proactive policies?
Automotive. Because of the Biden administration’s software restrictions and tariffs, we have massive restrictions on the import of Chinese-made electric vehicles into the United States. The way automakers are developing their EVs currently, they’re not making them as cheap or as well as some of the Chinese automakers. If Chinese cars get fully introduced into the United States, I worry we’re just going to lose even more of our automotive industry because people will just go to the cheaper option.