DJI explains why it won’t stop drones from flying over the White House — and what happens in a US ban

Time is not on DJI’s side.

The world’s largest drone maker has less than a year to convince the Trump administration to save its products from an automatic US ban — at a time when drone fears have been stoked more than ever.

Some of those fears were exacerbated by DJI’s own bad timing: days after a small DJI drone took out a plane fighting the LA wildfires, the company chose to announce it would no longer enforce the “no-fly zones” it designed to keep such things from happening. The company will no longer stop its drones from flying over airports, power plants, or even the White House, a move it says it planned months ago and already rolled out in Europe and the UK.

How does getting rid of no-fly zones make drones safer? What can DJI do to escape a US ban? And what might happen if that ban comes into effect?

We sat down with DJI head of global policy Adam Welsh and DJI public safety integration director Wayne Baker, and the answers seem to boil down to:

Below, you’ll find DJI’s detailed answers to our questions, condensed and edited for clarity.

How does removing DJI’s no-fly zones make drones safer?

Adam Welsh: First off, geofencing has been in place for more than 10 years, and we recognize any change to something that’s been in place for 10 years can come as a bit of a shock to people, so we recognize that your initial reaction and the reaction of some others is valid. We’re not trying to dismiss the concern. But let’s talk a little bit about why geofencing was first put in place and the limitations of that geofencing system.

We were the first people to put a drone in a box on a shelf that you could buy at Best Buy; anybody could go to a consumer electronic shop and literally in 30 minutes be up in the sky. Regulators were not prepared for that, and that’s not a ding against the regulator. That’s just [to say] that regulation takes time to formulate.

In the intervening 10 to 11 years, none of the regulatory agencies have chosen to mandate geofencing, and I think there’s good reason behind that. It was never going to stop a bad actor from doing something wrong. You could disable the GPS; you could wrap it in tin foil. That drone no longer knows where it is vis-à-vis the uploaded map onto the drone and you could fly wherever you wanted to.

To your point, it was a way of stopping new entrants from entering certain areas. But the regulators have basically decided to go a different route. They have not mandated geofencing. They have mandated for consumers to do community-based organization safety training. They have mandated LAANC permissions to access controlled airspace around airports, and they have mandated Remote ID. They have stuck to the basic principle that the operator should be in control of the drone, the airplane, or any other kind of aviation object at all times.

What we’ve done is we’ve switched to a system that means the operator has more accurate information but also maintains control of the drone at all times.

The FAA and regulatory bodies have used that language about the operator being in control of the drone at all times, but I’ve always taken it to mean there needs to be a pilot operating the craft who can see where the craft is so that it doesn’t hit something, not that the pilot should be able to freely go wherever they want. Would you disagree?

AW: But your Cessna [plane] does not prevent you from flying anywhere. It’s the pilot’s responsibility to know where they’re legally allowed to fly. So the same principle would apply to an operator of a drone. It is their legal responsibility to fly where they have been given permission.

Is it possible regulators decided not to mandate geofencing simply because DJI had already provided geofencing, and so the majority of drones already had it?

AW: I think saying that because the largest manufacturer has done something, there’s no reason to regulate, is not a good argument. You can’t say it’s fine that a drone made by a French or American manufacturer can fly where it wants with no safety issue and there’s only a safety issue if one of ours flies there. My suspicion is that when regulators have looked at this, they’ve realized how complicated it is.

To give you an example, we have done everything we can to try to smooth the process [for pilots to request access to a no-fly zone], yet it still requires that a human being look at every single application. You can imagine we process a huge volume of absolutely legitimate unlocking applications on a weekly and yearly basis. You would have an enormous burden on any regulator to go through that process.

Was that a big cost for DJI, and will there be a big cost savings by removing the hard geofences?

AW: I don’t think cost savings was really ever a focus when making the decision. Certainly the burden on our internal resources had been growing exponentially. We did beef up our resources considerably to make sure that we could try to do things 24/7, and we did our best to process within an hour of getting an application.

How do people benefit from the no-fly zones going away?

AW: I think somewhere around 50 percent of urban areas in the US are inside controlled airspace. So just say you’re doing a roofing survey. Half of your roofs require FAA permission to do a roofing survey. And that’s flying at very low altitude over a roof in a very short time period, so it’s intrinsically a very safe operation.

But if you got your FAA permission to do that, you might have a time block. You’re allowed to do it, but then you would have to wait for DJI hopefully to unlock within that time block. You might miss your timeframe. You might miss that job entirely. So the cost was quite considerable for our end users. But I think the one that really was a bit of a gut punch for us is that we frequently would hear from firefighters, first responders, and others who were trying to launch and maybe weren’t flying in their normal area.

They had gone into a disaster area to try and fly to assist a neighboring agency, and they were having to spend precious minutes or hours to get their drone unlocked. In an instance where somebody’s trying to launch a drone to save a life, we don’t want to impede that in any way, shape, or form.

We have had many instances where operators including first responders have not been able to fly a mission immediately when they need to because they have needed to get permission from us, even though they already have permission from LAANC or the FAA or whatever else to fly that operation.

Are there specific instances where DJI unlocking delays resulted in specific impacts to lifesaving work?

AW: We can’t actually give you the details of any individual unlock. There’s some privacy concerns around that.

Wayne Baker: While it’s not aviation safety that has been improved, our police and fire agencies would sometimes daily or weekly encounter areas that they needed to be able to fly in where we had restrictions. And they didn’t really have the time. An autistic child that’s missing in inclement weather — we didn’t have the time to go through “here’s our permissions” and all that. Several years ago, we tried to improve that process through the qualified entities program, and we still would have issues if pilots went to another area or changed logins. You just don’t have time.

I know it seems cliche when we say seconds and minutes count, but I can assure you they really do. I’ve had a lot of public safety reaching out about how this is good. They’re going to be able to have that confidence to take off when they need to.

Are there times where people have lost their lives or been injured, or property has been severely damaged, because DJI wasn’t able to unlock things in time?

WB: Unfortunately, we don’t have anything that says that, but you could reasonably infer that. One of the things I talk about is how great drones are at breaking a chain of events that could lead to an accident, an injury, or death, so you could reasonably infer that a delay could have reduced the chance that victim might have had [for survival] or that the building fire could have been mitigated sooner.

AW: We have so many documented instances of drones deescalating an active shooter situation or finding a child that’s on the verge of hypothermia or so many other things, that, for us, making sure that we’re not delaying to the point where one of those instances escalates into loss of life — that’s the key thing. We don’t have specific instances we can point to where there has been loss of life, but making sure that never happens is part of the reason behind this thinking.

Were any US regulators or representatives asking DJI to remove its no-fly zones?

AW: They didn’t directly ask us. We had some informal conversations with the FAA, and we’ve had informal conversations with aviation authorities in Europe and the UK, etc. Because this is always a voluntary measure, it’s something that DJI decided to do. The regulators have never really espoused an opinion one way or another. You chose to do it. Therefore, you can choose to remove it.

I think, primarily, the requests have been from government end users who’ve been restricted by this. Even from the very early days, we received numerous calls for us to remove this. An analogy: if you had a brand of car that prevented you from driving certain places after you got permission or restricted your speed in certain areas, I don’t think people would accept it. I think they would switch to a different brand or complain vociferously.

But you can’t drive onto a military base. You can’t drive onto the White House lawn. You can’t drive up to a nuclear power plant. There are fences. Why aren’t there geofences? Why not keep no-fly zones for the most sensitive of areas in the United States?

AW: We did look at this. We did consider whether we should keep certain specific areas. The decision was made internally that would still involve taking the onus away from the operator, and we wanted to stick to the principle that the regulators have approved, i.e., that it’s the operator’s discretion at the end of the day. To your point, an airplane can fly wherever it wants, but there are rules about where it can and can’t fly, and it’s up to the pilot to recognize that and live within it.

We’re doing a lot to make sure that our operators have information about where it is not appropriate to fly, giving them the warnings they need to make those decisions. But at the end of the day, the onus is with the operator.

WB: You can’t drive onto a military base because there are physical barriers those sites have put up, not the manufacturer of the vehicles. But that’s part of why we’ve looked at the maturity of the drone detection equipment out there. One example is with Axon’s Dedrone. There have been requests by the NYPD to Congress directly to allow them to mitigate drones as well as the Remote ID laws.

The DJI drone crashing into the Super Scooper plane fighting fires in Los Angeles — this was a drone that I do not believe would have been tracked by Remote ID as it’s too light. Is it a good time to let people fly their drones everywhere if Remote ID is not a good enough tool to stop drones?

AW: Just on the Remote ID front, we comply with the FAA regulations, and it’s part of the FAA regulations that, below 250 grams, Remote ID is not required. So that’s more a matter for the FAA to decide whether they want to change that. In terms of the flight restrictions around the fire area, there was a self-unlocking zone that was put in place over the fires. That was sort of our intermediate step.

The reason self-unlocking zones are pertinent is, quite often, firefighters need to fly their drones in those areas, but you also have to recognize that temporary flight restrictions are an even less reliable restriction than geofencing. There would never have been a permanent geofencing restriction in place over a fire, and to get that temporary flight restriction uploaded to a drone, that person has to connect to the internet, which automatically limits the percentage of people who would even get the TFR in the first place.

If you put a hard lock on that area, you could very well end up hampering the firefighters, but not hampering the guys that are rubbernecking. So I think the self-unlocking zone is the most appropriate approach to that kind of situation. It’s obviously regrettable that they didn’t stop this individual from flying in that area at the time.

[Editor’s note: At interview time, DJI wasn’t able to answer our questions about what was then an ongoing FBI investigation, but now that the investigation is over, the company tells us it did assist; it provided the email address and shipping information of the purchaser to an FBI agent. DJI says it doesn’t know whether the drone was broadcasting a Remote ID at the time.]

There is a countdown clock running. By December 2025, DJI needs the US government to designate an appropriate agency to determine whether it poses a national security risk, and to say out loud, no, we don’t believe it poses that risk, or else a complex process starts that keeps the FCC from authorizing your radios and cameras for import into the United States. How do you beat that clock?

AW: One thing to recognize is that we have been through security reviews in the past. The Department of Homeland Security reviewed our products back in 2019. They used Idaho National Laboratory, and we passed that review. The Department of Interior reviewed two of our products. Again, we passed that review. We’ve been reviewed even by the Pentagon, which has an office that looks at consumer off-the-shelf products. They looked at two of our products, and we passed that review as well. So we feel pretty confident in our technology and what we’ve done on data security.

You are 100 percent right: there is a ticking time clock. And that ticking time clock is even worse because of the political transition. It says 12 months from the signing. That puts you somewhere in December of 2025, but you’ve really lost the first two to three months while nominations take place and people are being put in place in the new administration. You’re then left with an eight- to nine-month time clock, and that’s where our concerns are. We’re absolutely going to embrace the scrutiny. We feel we have robust technology in place for data protection, and we actually are pleasantly surprised that we’re going to have an interlocutor.

There has to be some sort of designated party for us to have a conversation about where these risks are supposedly found and what kind of mitigations we could discuss if there are any things of concern. We want to embrace that. But to your point, what are we going to do now?

One piece is we feel it’s actually unfair to have the ticking time clock. Through no fault of our own, we could be put on the FCC’s covered list, which would prevent us from launching new products just because no agency took up the work. That’s our first concern that we would like to have conversations around.

There are about five agencies that are defined as appropriate [to determine whether DJI poses a national security risk]. It’s basically the Department of Defense, the FBI, the NSA, the Department of Homeland Security, and the Office of the Director of National Intelligence. It could be any one of those. All we ask is that the appropriate agency take up the work, do it within the required time period if that time period can’t be shifted, and that there be some sort of right of reply.

Just as your iPhone is continually getting firmware updates because of security patches, people will find flaws. There’s no such thing as a flawless piece of equipment. We just want to be able to have a conversation about what they deem a risk, if they find anything they deem a risk, and what mitigations we could put in place to deal with those potential concerns.

What is the actual work you do over the next three, six, nine, 12 months to get one of those agencies to take it up? Would DJI sue the government again?

AW: DJI doesn’t like to sue governments. That’s not how we like to spend our money. Our CEO likes to spend all his money on R&D making new product; he’s an electrical engineer and that’s where he likes to put his resources.

We have a government relations team, and we’ll be engaging both at the congressional level but also at the departmental level to try to make sure that this is taken up. We also know a lot of our end users are already getting in touch with us saying that they also want to make sure that this is done, and so I imagine they’ll also be requesting that the work is taken up.

We’re going to put our efforts behind trying to make sure it’s taken up through normal conversation, advocacy, and education. We’ll explain what we’ve done so far on data security, and we’ll explain the economic impact of DJI in the United States market. We’ve estimated that more than 400,000 jobs are being supported by the use of DJI equipment in the United States, and so there is a sizable impact to doing something negative on DJI. That puts the onus on the government, I think, to actually do this thoroughly within the timeframe.

What happens in the event of a ban?

AW: If we were added to the covered list, it’s not retroactive. What it would do is actually prevent us from launching new products because we wouldn’t be able to get the certifications we need for new products. So you’d be in kind of a perverse situation where the current Mavic model would continue to be on sale in the US, but a newer model would be available in Canada and Mexico and everywhere else in the world. We’d still be able to sell those models. Obviously, it’s not optimal for our end users and it’s not optimal for DJI as a company to only be able to sell what’s available as of 2025 in the US market.

How important is the US market to DJI, really? Is it so important that DJI would simply not be able to make or afford to make certain kinds of products?

AW: The US is clearly a significant market. Nobody should kid themselves that it’s not a significant market. Consumer spending and enterprise spending is hefty in the United States. But DJI has healthy markets across the globe, and that is more than enough to support the company and its research and development efforts.

The real tragedy around this is the collateral damage it would cause to resellers, which are American companies, and to the American software businesses that have built their entire businesses around DJI products. The biggest risk raised by venture capitalists is, “What’s going to happen with DJI?” It’s already hampering their ability to raise money and conduct their businesses. But for DJI at large, it would not stop us from continuing to grow.