Streetwise Professor

The evidence for SpaceX’s vital and unilateral role in (a) the Russian intercept of the ICEYE X36 satellite (which supports Ukrainian deep strikes into Russia) and (b) throttling the white-listed Starshield terminals Ukraine uses to defend against attacking Russian drone waves is massive, and nigh on indisputable.

When reports first broke that COSMOS 2612 had tracked and cut off ICEYE X36, R (aka @LibertyLynx) immediately suspected that the Russians did not have the tracking capability to pull that off on their own, suspected SpaceX involvement, and built an orbital model to test the theory. We then ran parallel models to cross-check the results, and the attached brief lays out the case in deep, and very technical, detail. It is based on the best and most complete publicly available information, including information produced by SpaceX, which is particularly damning

In brief, to recapitulate, on 16 October 2026 the Russian COSMOS 2612 sprinted from hundreds of kilometers away from ICEYE X36 and stopped on a dime within a few hundred meters of the US/NATO satellite. The likelihood that it could have done this on its own is near zero. Moreover, SpaceX was the only entity with the information and power to permit this intercept.

Maybe the brief will cause MEGO. But an animation (also created by R) that incorporates the information is an eye opener.

When I first viewed the video it reminded me of a football sweep play, with the Russian satellite in the role of a running back following the lead block of Starlink 35856, which pinned ICEYE X36 in place, allowing COSMOS 2612 to hit the hole right next to it.

Could have been a coincidence, right? Or the Russians just got lucky, eh?

Almost certainly not.

First, the Russians would not have had the information to guide COSMOS 2612 with that degree of accuracy, especially if ICEYE X36 had the ability to maneuver (more on that later because it is crucial). Russia has no space-based tracking assets, and relies on ground-based systems which are blind for 8-10 hours (due to the earth’s curvature). Further relying on COSMOS 2612’s onboard sensors limited range would have resulted in a miss of over 100km if ICEYE X36 had engaged in a brief avoidance burn.

But there was no burn. Why not? That is the central question here, the answer to which is particularly damning.

In response to an attempted approach in June, ICEYE X36 did engage in successful evasive maneuvers. Don’t believe me? How about ICEYE’s CEO, Rafał Modrzewski, who explicitly confirmed on the record to Reuters that ICEYE actively maneuvered its satellite to evade the Russian spacecraft, and that Russian maneuvers followed. Moreover, we have confirmed this in the data: independent reconstruction of all ~3,600 public Space-Track element sets explicitly verified X36’s maneuver on 25-26 June.

So why in June, but not in October? Because it couldn’t, that’s why. And why not? Because it was pinned. By whom, and how? By SpaceX, exploiting its unprecedented control in space, courtesy in large part to FCC regulations.

In brief, SpaceX issued a collision alert to ICEYE X36 for Starlink 35856, based on an assertion that ICEYE X36 would be within .401 kilometers of Starlink 35856. Under FCC Condition 16–negotiated with SpaceX lawyers–this collision alert obligated ICEYE X36 to eschew any maneuvers. In other words, the alert fixed it in place, thereby enabling the COSMOS 2612 to engage in a braking maneuver that put it right beside the American satellite.

Damning fact #1: SpaceX’s own documentation demonstrates that the collision alert was unwarranted and unnecessary, because in fact the distance to Starlink 35856 was 4.8 to 5.0 times greater than the claimed .401 km. Based on the correct distance, no alert was needed.

Damning fact #2. Starlink satellites are known for their autonomous collision avoidance abilities. If there was a genuine .401-meter collision risk, STARLINK-35856 could have—and normally would have—autonomously pulsed its thrusters to clear the corridor. Instead, SpaceX withheld that autonomous maneuver. By choosing not to move, they intentionally shifted the legal burden entirely onto ICEYE via FCC Condition 16. (Why should SpaceX insist that a vital US intelligence satellite yield right of way to one of its many satellites that primarily provide prosaic Internet connectivity?)

In brief, SpaceX engaged in acts of commission and omission without which the intercept would have been impossible.

To add to the disturbing implications, consider the following. The Russians would have known that it would have been futile to attempt an intercept if-as had been demonstrated in June-ICEYE X36 would be free to maneuver. The sprint and braking maneuvers were costly, and it would have been irrational to engage in them if the probability of success was nil. The only way the probability of success would have been sufficiently high to justify the cost would have been if the Russians were highly confident that ICEYE X36 would be pinned. And how could they have known that?

The plausibility of coincidence or luck vanishes when one considers the fact that the intercept was contemporaneous with the throttling of Ukraine’s Starshield terminals. These were whitelisted terminals, used to guide Ukrainian drone defenses. Once a Ukrainian drone exceeded the speed limit imposed by SpaceX, its connection was dropped, making it stupid. Note that Ukrainian drones operating within Russia were already limited to much lower speeds.

And by the way, when provided this information, Grok agreed that there was no coincidence here comrade, and assigned a less than .01 percent probability to this being a random coincidence.

Musk’s shifting stories also belie any innocent explanations. At first he said that the decision to throttle Ukrainian drones was a WH decision, because Starshield is operated by the US government. But that explanation exploded later when Hegseth and Rubio both stated that SpaceX had acted unilaterally pursuant to its Terms of Service with the US government. Musk later embraced the ToS story, claiming that they prevented Ukraine from using Starshield to kill anyone, and that Ukraine had violated this provision.

They just started doing that in October, Elon? Were you shocked! Shocked! to learn that Ukraine was using Starshield/Starlink for military purposes?

Interestingly, among the people whom Musk claimed he was protecting were “poor Russian conscripts.” Putting aside the fact that Russian troops are not conscripts (although they are often shanghaied into contract service), said Russian troops are killing Ukrainians. And further, by disabling the terminals, Musk increased the lethality of Russian strikes against Ukraine and Ukrainian civilians in particular.

Interesting to know Elon’s solution to this particular trolley problem.

The main point here is that SpaceX played a central role in compromising both Ukraine’s defensive and offensive capabilities at the precise same time. And in explaining the former, he engaged in a virulent rant against Zelenskyy. In that context, any innocent explanation of the intercept crumbles.

The implications of this go well beyond the events in Ukraine, as the brief discusses in detail. Unlike legacy defense contractors, including those involved with highly classified space and satellite operations, SpaceX operates on a “CO-CO” model: contractor owned, contractor operated. This contrasts with the legacy GO-GO model, in which the government buys, owns, and operates what the contractors produce.

The contractor-SpaceX here-provides services according to a contract: the “Terms of Service” that appear so notoriously in the foregoing. It is clear that these ToS provide tremendous discretion to SpaceX to make decisions that have grave national security implications. It is also clear that these it is willing to exercise these powers in ways that are inimical to US security interests.

And don’t think this is the only time. During the Iran War SpaceX invoked its ToS to prevent the use of Starlink to mount LUCAS drone strikes against Iran. Until, that is, the Pentagon agreed to pay a 4x greater fee to SpaceX. Which reminds me of the old Churchill line: Now we know what you are. We are just haggling over price.

This is all the more disturbing given the greatly expanded role of SpaceX, and Grok in particular, in every aspect of US military operations. It is sickly ironic that Hegseth and his tech chief Emil Michael declared Anthropic a national security risk for demanding that it retain a veto over the use of Claude given that they have supplanted it with a contractor who has used such a veto to the detriment of US security interests.

In the present instance, how bizarre is it that the US military would enter into a contract which allowed its counterparty to veto decisions to kill people? After all, the military’s job is, well, to kill people. The decision to do so is vested in Constitutionally established authorities operating subject to US law. It is not vested in Elon Musk, and it is incomprehensible that the US government has entered into contracts that grant him that authority.

This is a classic problem in transaction cost economics. SpaceX provides very specific assets (specific in the technical economic sense) to the United States government, and the Department of Defense in particular. The specificity of these assets gives SpaceX incredible power to hold up the government. Contracts should be crafted to limit this holdup power. Instead the Pentagon has enabled it.

GOGO is the default organizational response to holdup problems. Given the acute severity of the holdup problems posed by SpaceX, unless the government can find a way to write much better contracts with it, it should move back to that model.

But hot-from-the-presses news suggests that this is a vain hope, at least under the Hegseth Pentagon. The Defense Intelligence Agency just released “Challenges to Security in Space: Congestion, Competition, and Cooperation.” It relies on a Russian propaganda site (topwar.ru) and TASS (!) to hype Russian capabilities (namely its ground-based Krona surveillance complex).

The reliance of this document on a trash Russian military site renders it little better than the Steele Dossier, or the hack job on Russian election interference that the outgoing Obama administration cobbled together in January 2017.

Moreover, the timing is rather, shall we say, odd. Lo and behold, the day after Elon offered his fourth explanation for throttling Ukrainian drones, specifically that if he didn’t Russia would target Starlink satellites, the DIA drops a report citing topwar.ru to warn of Russian threats to American satellites. A coincidence, surely. If you believe that . . .

But there are indications that American dogs are fighting under the carpet–specifically the State Department and the Pentagon. Rubio was clearly displeased at the Ukraine drone throttling, and the fact that this was a Musk decision. The State Department was also infuriated by the intercept, taking a complaint to the UN. And Rubio just recently fired one of Musk’s operatives. So we must hold out hope that Rubio (whom Trump surely holds in greater esteem than Hegseth, and hopefully Musk) will be able to put a stop to this madness.