Opinion | Elon Musk’s control over satellite internet demands a reckoning

Starlink is far from the first constellation of satellites in low Earth orbit and far from the first to sell to militaries. But what distinguishes the network is the amount of data it can move, as well as how quickly it can increase that capacity: SpaceX can launch satellites unprecedentedly fast and at unprecedentedly low cost thanks to the reusable rockets it has pioneered. The bigger the satellite fleet, the more versatile and effective: As a satellite flies above a terminal located on the ground, it transfers the signal to a satellite behind it, and so on, forming a chain that ensures users maintain constant access to the internet.

Working together, these satellites can provide this kind of service even in remote locations. And, of course, they can provide it in locations where traditional methods of connection, for whatever reason, are unavailable — such as war zones.

Starlink has been a boon to Ukraine’s efforts to repel the Russian invasion ever since Ukraine’s digital minister tweeted a plea for assistance at Mr. Musk in February 2022 and Mr. Musk answered. The trouble with relying on one key technology for battlefield communications, however, is that it also means relying on whoever controls that technology. And because Starlink is a commercial company rather than a traditional defense contractor, in this case Mr. Musk has largely been able to call the shots — literally.

Using a technique known as geofencing, Mr. Musk has restricted his satellites’ availability on the front lines. Purportedly, he’s wary of aiding an offensive rather than merely a defensive war, which isn’t surprising given his proposal for a “peace plan” whose defining feature was its generosity toward Russia. Sometimes, when Ukraine has wanted to retake Russian-occupied cities, its government has had to message SpaceX employees or the CEO himself for permission. Ukraine’s request at one point to send a drone into ships docked in the Black Sea near Crimea was reportedly rebuffed.

Mr. Musk is the wrong person to be making these calls. The Defense Department announced a deal with SpaceX in June. (Ukraine had been getting Starlink capacity for free until the richest man in the world, not unreasonably, changed his mind.) The details aren’t public, but reporting suggests that the Pentagon will determine when and where the 400 to 500 new devices it is purchasing will work — within Ukraine’s territory, at least. The short-term solution to the Starlink conundrum is for the United States to continue this kind of negotiation, allowing itself more autonomy to assist Ukraine in carrying out crucial missions.

The longer term, however, looks complicated. Ukraine is unlikely to be the only country friendly to the United States that ends up in a conflict with a foe eager to sabotage its internet. The region attracting the most apprehension at the moment, unsurprisingly, is East Asia, where Taiwanese officials are mulling how to handle the possibility of a Chinese invasion in which the undersea cables that connect the island to the world’s communications infrastructure are cut off. Given that the success of Tesla, the electric-car company that Mr. Musk co-founded and leads, depends on its ability to produce cars in China, counting on Starlink seems a poor idea. It might be a poor idea for the U.S. government more generally, too, if the United States wants to be able to choose how it defends its own interests and its allies’.

What’s to be done? While a president theoretically has the legal option of nationalizing Starlink in a worst-case scenario, as Woodrow Wilson did with the country’s railroads during World War I, that would be neither politically popular nor prudent. A better solution might be for the United States to try to build satellites of its own. The $1.5 billion contract the Pentagon awarded last week to Lockheed Martin and Northrop Grumman to create a low-orbit satellite constellation is the start of such a strategy.

The plan aims for only 72 satellites, roughly 1 percent of the number SpaceX has put into the sky. All the United States has to do, however, is muster a fleet large enough to enable smaller-scale but essential operations, so Starlink is no longer the democratic world’s only good option. This might not happen quickly, given the slowness and scleroticism for which the giants of the realm of military contracting are known. Encouraging Silicon Valley hotshots and others in private industry to inject some competition might help. The upside could be less waste and more innovation — exactly as SpaceX and Starlink have demonstrated. The downside is just as obvious.