How Musk’s Starlink became a security liability for the U.S.

Here on Earth, the satellites that make up Starlink look like a string of stars traveling across the night sky. More than 4,000 of them are circling Earth in low orbit right now. They’re part of the private venture that’s the brainchild of billionaire and SpaceX founder Elon Musk.

Last year, when Russia invaded Ukraine, Musk sent Starlink terminals there so Ukraine could stay connected to the internet. But it turns out Musk controls both the on and the off switch on that technology, giving him an outsized role in the conflict, says Steven Feldstein of the Carnegie Endowment for International Peace.

He’s written a story in The Atlantic on how that happened and what can be done about it. The following is an edited transcript of his conversation with Marketplace’s Lily Jamali.

Steven Feldstein: One thing that Elon Musk was able to do was, he was able to get a lot of these low-Earth orbiting [or LEO] satellites up into space very quickly because of the twinning with SpaceX rockets. And what we also know is that the more satellites you have in place, the better your connectivity is when it comes to having greater bandwidth and lower latency for access. And so the more that Elan Musk is able to do that, the more he is able to shoot up hundreds of Starlink satellites every month, the greater the advantage that Starlink has over any potential competitor.

Lily Jamali: I hear what you’re saying about scale. But I was really surprised to read that statement from you about how no other company and no government can match this capability that Starlink provides right now. I mean, really? The government isn’t capable of bringing this technology to bear, or at the very least funding competitors?

Feldstein: It would take a long amount of time. And first of all, my understanding is there’s a technological component. They’ve been able to put together advanced software to make this work. And it’s something that takes a lot of trial and error. I think the second thing is that just being able to put that many satellites into orbit all at once is not something that would come easy. It would be costly. And so I think those two aspects present a pretty high bar when it comes to being able to all of a sudden have a rival that would have a comparable service that could be used. It certainly can happen. There’s no reason why, technologically, it can’t happen. But there are significant financial and logistical obstacles to doing that anytime in the near future. That, I think, is the big problem.

Jamali: Yeah, and it sort of strikes at this other issue, which is that the U.S. government has outsourced much of the space program to SpaceX, and Starlink is a subsidiary of SpaceX.

Feldstein: That’s correct. I mean, this just goes back to a larger trend of commercialization that has happened in the post-Cold War era, where a decision has been made by the U.S. government to really work with a web of private companies, many of which have other motives and other commercial interests. And so the more you commercialize different sectors, the more these companies have other clients. And essentially, they can say, “Well, U.S. government, you need this, but we also have other clients here, and how we’re going to prioritize that and what our interests are might diverge.”

Jamali: So are there any existing competitors at all, even in the broadband market, which Starlink is a part of?

Feldstein: There are some, but my understanding is that they’re pretty fledgling. And I think one of the bigger problems, especially when it comes to LEO satellites, is no one else has really been able to match what Starlink has done. Now, there are some competitors like Amazon that are interested in doing more with that, but they’re pretty far behind when it comes to being able to rival where Starlink currently is.

Jamali: From a technological standpoint?

Feldstein: Yeah, I mean, my understanding is that even as recently as a few months ago, they attempted to put up their very first two satellites, but ran into rocket launch problems and had to scrap it. So that tells you something. In the meantime, while they’re struggling to get two LEOs up, Starlink had put up another several hundred. So there’s a pretty large discrepancy between the two at the moment.

Jamali: You write about how the Defense Production Act is being floated by some to deal with this situation. What are experts telling you about how that might be deployed in this scenario?

Feldstein: The DPA, as it’s referred to, is something that actually is invoked quite often. So it’s actually a relatively straightforward contractual process to use. I think it’s something that many people think would be viable in the long term, particularly if trying to come up with a regular, global contract was something that failed. But it’s still something that would be another step. And in the meantime, coming up with contractual arrangements, which is something that Pentagon actually has just started to do as of June, it represents a more straightforward path.

Jamali: And could a government takeover of Starlink be viable? And would that even be desirable, do you think?

Feldstein: Yeah, so that’s an idea I explored in the piece. And there is precedent for doing so. But usually, that happens after either a national security crisis or financial crisis. So think about the creation of the TSA [Transportation Security Administration] after the 9/11 attacks. So I’m not sure that we’re at a moment where there’s a crisis that would precipitate that. I think the other question would just be, would a government-run entity be as innovative as what we’ve seen with Starlink? I mean, the good thing with Starlink and with Elon Musk’s leadership is that he has been tremendously creative in terms of pushing the boundaries of what people thought was possible with satellite internet technology. The downside is that Elon Musk is an unreliable actor, and has made lots of impulsive decisions that undermines the credibility of the service. I think in general, you want to maintain the innovation, so nationalization to me seems like a very messy route forward.

Jamali: Until this week, do you think it was clear just how enmeshed Elon Musk was in U.S. foreign policy, even to people who study this for a living like you?

Feldstein: I think there has been a kind of growing awareness, in large part because of the Ukraine war. And I think the more the reports come out, the more we know from the inside just how fragile that relationship is and just how concerned the Pentagon is. It doesn’t have control over a critical resource that is absolutely vital to the success of Ukrainians and their counteroffensive and beyond.

Jamali: The title of your article is “The Answer to Starlink Is More Starlinks.” Tell me what you mean by that.

Feldstein: I think the issue is that we can come up with contractual arrangements with Starlink under Elon Musk’s leadership, and that will help provide more reliability in the short term. But the longer-term problem is that as long as you have a figure like Elon Musk at the helm of a critical national security asset, you’re never going to solve this reliability problem. And so the best way to get around that would be to have market diversification, to come up with alternatives and other companies, competitors that would allow greater options for the U.S. military when it comes to a critical service. And so that to me is ultimately the kind of bottom line — is that in the short term, look for contractual arrangements to try to box in and bring about more reliability; in the long term, explore options for market diversification so you’re not reliant on a single operator and a single individual like Musk to provide this capability.

Jamali: And so if you’re the secretary of defense, how do you make that happen?

Feldstein: Well, money is a good start, and thinking about how you would apply those funds and in what way. For example, what kind of rockets would be available? If you can’t use SpaceX rockets, can you use others? Can you apply pressure so that you can use different types of legislation to ensure that the right types of rockets to bring these satellites into orbit are available? These are the types of questions I think ought to be explored — but again, in a longer-term fashion. In the immediate term, when it comes to Ukraine, when it comes to other potential military conflicts, there needs to be a way to bring about greater resolution to what really is an untenable situation when it comes to Starlink and Musk’s discretion.

More on this

Concerns over Musk’s control over Starlink aren’t limited to Ukraine. Musk, of course, also runs electric carmaker Tesla, which has a giant factory and market presence in China. Steven Feldstein writes in The Atlantic about Musk’s claim that the government there has pressured him to keep Starlink off-limits to Chinese citizens.

This can all feel a little abstract. So if you want to see what Starlink satellites look like in real life, here are some images captured by National Geographic photographer Babak Tafreshi from national parks in California, and a New York Times rendering of Starlink’s network in Ukraine.