Starlink, Ukraine and the Invisible Hand of American Power: Who Has Been Turning Off the Lights?

By Matthew Parish

Thursday 15 October 2026

There is something peculiarly disturbing about the proposition that a modern army, fighting for the survival of its country against a vastly larger aggressor, might discover in the middle of military operations that an essential component of its communications infrastructure has been restricted by somebody sitting behind a computer terminal thousands of miles away. The soldiers involved may have spent months preparing their equipment, perfecting their operational procedures and identifying their targets. Their commanders may have invested considerable resources in sophisticated technologies intended to overcome the numerical superiority of their adversary. Yet all these efforts may be frustrated by a modification to a satellite communications service over which neither the soldiers nor their government exercise ultimate control. The events reported in Ukraine on 14 October 2026 suggest that this is no longer merely a theoretical possibility. They raise questions about the relationship between private technological power, American foreign policy and Ukrainian national sovereignty that reach considerably beyond the immediate military consequences of the incident.

The first question is whether the reported restrictions on Starlink communications are genuine. On the available evidence, there is substantial reason to believe that they are. The second question is who authorised them, and here matters become considerably more complicated. The third, and perhaps most intriguing, question is whether these restrictions were imposed for technical or contractual reasons, or whether they represent an exercise in political coercion at a particularly sensitive moment in the relationship between Washington and Kyiv. It would be irresponsible to assert that we possess definitive answers to the latter questions. Nevertheless the available evidence, combined with the history of Starlink’s involvement in Ukraine and the contemporary diplomatic circumstances, permits a reasonably detailed examination of the possibilities.

On the night of 13–14 October 2026, Ukrainian military personnel began reporting unexpected interruptions in Starlink connectivity affecting drones travelling at substantial speeds. Serhii Beskrestnov, an adviser to President Volodymyr Zelenskyy on defence technology, confirmed that restrictions had been imposed on the military version of the service, known as Starshield. Initially the problem appeared to arise when terminals were travelling faster than approximately 160 kilometres per hour. Subsequent reports indicated that connections were sometimes interrupted at substantially lower speeds, with some military units recording difficulties between approximately 72 and 120 kilometres per hour. Ukraine’s 7th Air Assault Corps reported that drone connections were disappearing for periods of 10 to 20 seconds, and that several operations had been cancelled because their continuation had become impracticable. These were not merely complaints circulating amongst disgruntled social media commentators. Ukrainian Defence Ministry sources confirmed the restrictions to several news organisations, while individual military commanders provided accounts of the operational consequences.

This distinction between an interruption in service and a deliberate restriction is important. A satellite communications system is vulnerable to various technical difficulties, including atmospheric interference, software faults, electronic warfare and congestion. Russian forces have devoted considerable effort to disrupting Ukrainian satellite communications, and occasional failures are an unavoidable feature of modern military technology. Yet the reported interruptions exhibited a particular characteristic: they were associated with the speed at which the receiving equipment was moving. That is consistent with the application of a mobility restriction, rather than a general satellite failure. It does not establish why the restriction was introduced, nor does it establish that Ukrainian military operations were its intended target. Nevertheless, the accounts from Ukrainian officials point towards a change in the conditions under which the service was being provided.

It is also necessary to distinguish Starlink from Starshield. Ordinary Starlink is a commercial satellite internet service, employed by civilians and military personnel alike, while Starshield is designed for governmental and national security applications. According to Ukrainian military sources, ordinary Starlink terminals continued to operate, whereas the higher-speed military service encountered difficulties. The available reports therefore do not establish that Ukraine’s entire military communications system was disabled. They establish something narrower, but still militarily significant: certain sophisticated drone operations, particularly those involving movements at speed, became more difficult or impossible. The Polish Centre for Eastern Studies, in an analysis published on 14 October, also cautioned against assuming that the restrictions had comprehensively undermined Ukraine’s defences against Russian attack drones. It observed that many interceptor systems employ ordinary Starlink rather than Starshield. The available evidence thus supports the existence of a serious operational restriction, without justifying every claim made about its consequences.    

The significance of this development becomes clearer when we consider the evolution of Ukrainian military technology since February 2022. Starlink initially provided an invaluable means of maintaining communications when conventional telecommunications infrastructure was being destroyed or disrupted. It enabled dispersed military units to exchange information, transmit reconnaissance imagery, coordinate artillery and communicate with headquarters without depending upon vulnerable terrestrial networks. As the war developed, Ukrainian engineers and military innovators found increasingly sophisticated applications for satellite communications. Amongst the most consequential was their incorporation into remotely controlled aerial and maritime drones, permitting operators to maintain communications at considerable distances from the front line. Such systems have enabled Ukraine to strike Russian logistics, military installations and other targets in occupied territory, helping to compensate for her deficiencies in conventional military resources. The value of the technology has therefore expanded from maintaining communications to enabling important categories of military operation.

This development contains an uncomfortable paradox. One of Ukraine’s greatest military advantages has been her capacity to improvise, employing commercially available technologies in ways that their manufacturers did not necessarily envisage. Yet commercial technology remains subject to the contractual arrangements, technical restrictions and political circumstances governing its providers. The more indispensable a particular system becomes, the greater the leverage enjoyed by whoever controls it. In the nineteenth century, governments understood that the security of international telegraph cables might determine their ability to conduct war or diplomacy. In the twentieth century, they appreciated that control over telecommunications networks, radio frequencies and eventually satellites constituted an essential attribute of national security. In the twenty-first century, some of these functions have migrated into the hands of corporations whose commercial and political interests do not necessarily coincide with those of the governments depending upon them.

The obvious person to consider in connection with the present restrictions is Elon Musk, the founder and chief executive of SpaceX. His company owns and operates the underlying satellite technology, and he has previously demonstrated both a personal interest in the political consequences of the war and an extraordinary degree of influence over the conditions under which Ukraine receives satellite communications. There is also a substantial historical precedent for considering the possibility of his personal involvement. In July 2025 Reuters published a detailed investigation alleging that Musk had personally ordered restrictions on Starlink coverage during Ukraine’s counteroffensive in the Kherson region in September 2022. According to several sources familiar with the episode, Musk instructed a senior SpaceX engineer to restrict connectivity in areas where Ukrainian forces were attempting to recover territory from Russian occupation. The resulting interruption reportedly disrupted reconnaissance, artillery operations and efforts to surround Russian positions. SpaceX disputed the accuracy of the Reuters investigation without specifying precisely which elements of the account it considered incorrect.

That investigation deserves particular attention because it goes substantially further than the familiar controversy over Starlink’s availability around Crimea in 2022. In the Crimean episode, Musk refused a Ukrainian request to extend satellite coverage to support an operation against Russian naval vessels. Earlier reports suggesting that he had actively switched off an existing connection were subsequently corrected following a clarification by his biographer Walter Isaacson. The Kherson allegations, by contrast, involved an affirmative order to withdraw coverage that Ukrainian forces had previously been using. Reuters reported that the decision may have reflected Musk’s concerns about Russian nuclear retaliation during the Ukrainian counteroffensive. Those concerns were not peculiar to Musk: some American officials were also contemplating the possibility that battlefield reverses might persuade Moscow to escalate the conflict. Nevertheless, the distinction between refusing to provide additional assistance and actively removing existing operational capabilities is an important one. It illustrates how decisions apparently relating to the administration of a commercial service may have consequences equivalent to decisions traditionally reserved to military commanders and governments.   

There is also a further precedent involving SpaceX’s corporate management. In February 2023 Gwynne Shotwell, the company’s president, acknowledged that SpaceX had taken measures to prevent Ukraine from employing Starlink for certain offensive military applications. Her explanation rested upon a distinction between supplying communications technology to support Ukraine and permitting its use as an integral component of weapons systems. That distinction has become increasingly difficult to maintain as Ukrainian military innovation has progressed. A satellite terminal transmitting battlefield information may be supporting defensive communications at one moment and directing an unmanned aircraft towards an enemy position at another. The same technology might assist with delivering essential supplies to an isolated Ukrainian military position or destroying a Russian ammunition depot. Consequently, restrictions intended to distinguish permitted from prohibited applications can have much broader effects than their authors anticipate.

These historical incidents make the possibility of a decision within SpaceX worthy of serious investigation. They do not, however, establish that Musk personally ordered the restrictions of October 2026. Indeed, there is an important complication that has received insufficient attention in some of the public discussion. In March 2026, Musk stated that ordinary Starlink was not licensed for use as part of weapons systems, but that Starshield was a separate network operated by the United States government and not under SpaceX’s control. This is a significant assertion. If accurate in the relevant contractual and operational sense, it suggests that the restrictions imposed upon Ukrainian military users might have originated within the American government, rather than with Musk personally. SpaceX might have implemented a technical modification, while the decision to impose it could have been made elsewhere. On the other hand, the precise allocation of authority between the company and the American government has not been publicly established for the Ukrainian arrangements. A statement that a government operates a particular service does not necessarily mean that the manufacturer has surrendered every means of modifying its technical configuration.

We are therefore confronted with two potentially overlapping centres of authority. The first is SpaceX itself, where senior management and technical personnel possess knowledge of the system and the capacity to implement changes. The second is the United States government, whose contractual arrangements with SpaceX may permit authorised officials to determine the conditions under which Starshield is made available to foreign military users. If the October restrictions resulted from a deliberate political instruction, responsibility would have to be traced through one of these channels, or through cooperation between them. It would be premature to identify Musk, Shotwell or any particular American official as the person who necessarily authorised the change. A routine technical or contractual modification might even have been introduced without senior political involvement, although that would raise questions of its own about the safeguards governing changes to an active wartime communications system.

Nevertheless, if we consider the hypothesis that the restrictions were politically motivated, the contemporary circumstances immediately direct our attention towards President Donald Trump and his administration. During the days preceding the interruptions, Washington had been increasing pressure upon Ukraine to suspend attacks against Russian oil infrastructure. Reports published on 11 and 12 October indicated that American officials had threatened restrictions upon intelligence cooperation and military assistance unless Kyiv agreed to halt strikes against Russian refineries. The immediate context was a dramatic deterioration in relations between Trump and Zelenskyy, accompanied by the American President’s controversial agreement to relax restrictions on Russian diesel exports. Trump attributed part of the increase in American fuel prices to Ukrainian attacks upon Russian petroleum facilities, although the wider disruption of energy supplies associated with the conflict involving Iran and the Strait of Hormuz provides a substantial alternative explanation.    

The political calculations surrounding energy prices are readily comprehensible without recourse to elaborate conspiracy theories. An American President approaching midterm congressional elections has a powerful interest in controlling the domestic economic consequences of rising fuel costs. The ongoing conflict involving Iran has complicated the international energy market, and Ukrainian attacks upon Russian oil infrastructure introduce an additional source of uncertainty. From Washington’s perspective, therefore, pressing Ukraine to suspend those attacks might appear to offer a means of easing pressure upon global energy supplies. From Kyiv’s perspective, however, the attacks are an important component of her military strategy, designed to impair Russia’s capacity to finance, supply and sustain her invasion. The disagreement is not simply about the merits of one particular category of military target. It reflects a more fundamental divergence between the immediate economic priorities of the American administration and Ukraine’s longer-term military requirements.

There is also a possible Russian dimension to the American calculations. President Vladimir Putin has an obvious interest in bringing Ukrainian attacks against Russian oil installations to an end, particularly insofar as those attacks threaten military logistics and economic stability. Trump’s recent agreement concerning Russian diesel and his announcement of an energy ceasefire created a diplomatic framework within which Moscow could seek relief from those attacks. Yet Russia continued striking Ukrainian energy infrastructure even as Ukraine indicated her willingness to observe a temporary suspension of attacks upon Russian refineries. The asymmetry is consequential. If Ukraine refrains from attacking Russian energy installations while Russia continues destroying Ukrainian electricity generation and distribution systems, the practical benefits of restraint are distributed very unevenly. It does not follow that Putin requested restrictions on Starshield or influenced a particular technical decision, and no reliable public evidence presently establishes such a connection. Nevertheless, the strategic advantage Russia would obtain from restrictions upon Ukrainian drone operations is not difficult to identify.   

The question of Musk’s personal political position adds another element to this analysis. On 10 October, Musk publicly endorsed Trump’s criticism of Ukrainian attacks against Russian oil installations, writing that Zelenskyy had gone too far. This was not an isolated comment by an entrepreneur entirely indifferent to foreign affairs. Musk has repeatedly engaged in public discussions concerning the terms upon which the war might be concluded, and his earlier interventions have sometimes provoked criticism in Ukraine for appearing to accommodate Russian territorial demands. His political relationship with Trump has experienced periods of public disagreement, but their positions concerning Ukrainian oil strikes were demonstrably aligned in the days immediately preceding the Starshield restrictions. It is therefore reasonable to investigate whether that alignment had any practical consequences for SpaceX’s operational decisions. What cannot be established merely from a public statement, however, is that Musk gave an instruction to restrict communications, or that the White House requested him to do so.   

We should consequently distinguish three possible explanations for the events of 14 October. The first is that SpaceX introduced a technical or contractual restriction for reasons unconnected with the diplomatic dispute. This possibility cannot simply be dismissed. In February 2026, Ukraine and SpaceX cooperated to introduce a registration system designed to prevent Russian forces from employing unauthorised Starlink terminals. That undertaking involved technical restrictions upon equipment operating in Ukrainian territory, because Russian personnel had acquired terminals and incorporated them into military systems, including drones. It is conceivable that subsequent adjustments to service permissions, mobility limits or security arrangements might have caused unintended consequences for legitimate Ukrainian military users. There is also the possibility of a configuration error or an inadequately communicated change to a particular class of subscription. Such explanations would not eliminate responsibility for the disruption, but they would fundamentally alter its political meaning.    

The second explanation is that senior figures within SpaceX independently decided to restrict the military applications of Starshield because they were concerned about the escalation of the conflict or the deployment of their technology in particular categories of attack. There is some historical support for considering this possibility, given the company’s earlier statements and the reporting concerning Musk’s conduct in 2022. A decision of this nature might have been influenced by fears of Russian retaliation, American regulatory concerns, commercial relationships or the general difficulties confronting a corporation whose technology has become embedded within active military operations. Yet the present restrictions reportedly affect Ukrainian operations within occupied territory, not merely long-range strikes against Russia’s internationally recognised territory. The breadth of the resulting disruption would therefore require explanation. If the intention were to prevent a narrowly defined class of offensive operations, a restriction interfering with multiple categories of military activity would appear to be a relatively indiscriminate means of achieving it.

The third explanation is that the restrictions constitute a form of pressure connected with the policies of the Trump administration. Under this hypothesis, an American governmental authority might have directed or authorised a modification to Starshield’s operating conditions as a warning to Kyiv that continued access to particular military technologies depended upon Ukrainian compliance with American diplomatic preferences. Alternatively, SpaceX might have modified its service in anticipation of Washington’s wishes, without receiving an explicit presidential instruction. Such behaviour is not inconceivable in an environment in which the company maintains extensive contractual relationships with the American government and is sensitive to the political circumstances governing those relationships. However, there remains a considerable evidential distance between identifying incentives for compliance and demonstrating that a particular decision was taken in response to those incentives.

There is a further difficulty with the political hypothesis, and it deserves emphasis. Reuters reported that Ukrainian drones used to strike Russian oil installations rely upon communications arrangements other than Starlink. Consequently, restricting Starshield would not necessarily prevent the very attacks that Washington was seeking to discourage. Its most immediate effect appears instead to have fallen upon the medium-range drone operations, reconnaissance activities and logistical capabilities of Ukrainian forces. Some of the reported affected missions were connected with the defence of Ukrainian territory rather than attacks upon Russian oil facilities. If the intention was political coercion, the measure may therefore have been designed less to disable a specific military operation than to demonstrate the consequences of American displeasure. It is equally possible, of course, that the mismatch between the reported political objective and the military consequences indicates that no such objective motivated the restrictions in the first place.   

The simultaneous controversy surrounding American intelligence sharing has intensified Ukrainian suspicions. President Zelenskyy complained that Ukraine had not received the usual advance information concerning Russian ballistic missile attacks upon Kyiv on consecutive nights. These attacks caused substantial civilian casualties and damaged important energy infrastructure, leaving hundreds of thousands of residents without electricity. It is understandable that Ukrainian officials should interpret failures in advance warning, occurring alongside restrictions upon Starshield, as possible elements of a broader deterioration in American support. Yet here too the evidence requires careful examination. Trump denied that intelligence sharing had been suspended, and Ukrainian officials acknowledged that some warnings continued to arrive. The Polish Centre for Eastern Studies noted that failures to detect or communicate particular missile launches had occurred previously for technical reasons. Coincidence in timing may raise legitimate investigative questions, but it does not prove that two separate systems were deliberately restricted pursuant to a coordinated political policy.    

We should also be cautious about the temptation to personalise every decision within a vast technological or governmental organisation. Elon Musk is an exceptionally prominent figure, and his public conduct makes him an obvious subject of scrutiny. Donald Trump, likewise, occupies a position from which he can influence the conditions governing American military assistance. Yet the administration of Starshield involves technical engineers, contractual authorities, military departments and potentially classified arrangements between governments. It would be entirely possible for a decision originating within one part of this institutional structure to be implemented by another, without every participant understanding its full political or military implications. Equally, a configuration change might be authorised by a relatively junior official within parameters established by more senior figures. Responsibility must therefore be examined at several levels: who initiated the change, who approved it, who implemented it, who knew that it would disrupt Ukrainian operations, and who possessed the authority to reverse it.

These questions should be susceptible to documentary investigation, even if the relevant materials are not presently available to the public. SpaceX should possess records indicating when the relevant service parameters were modified and under whose authority. The contractual arrangements governing Starshield ought to identify the respective powers of the company and its governmental customers to impose technical restrictions. American officials should be capable of determining whether the change was authorised by a department, agency or military command, or whether it originated independently within the company. One particularly important question is whether similar restrictions were imposed upon other Starshield users outside Ukraine. Another is whether the affected Ukrainian terminals belonged to a particular contractual category whose operating conditions had been altered. Answers to these questions would help distinguish a targeted political intervention from a broader technical or administrative change. They would also identify the institutional safeguards, if any, intended to prevent unexpected disruptions to military operations.

The international implications are considerable. Poland has financed a substantial proportion of the Starlink equipment and service used by Ukraine, yet Warsaw does not necessarily possess the technical authority to guarantee the continued availability of the communications for which it pays. Polish Foreign Minister Radosław Sikorski responded to the October restrictions by publicly questioning Musk about the possibility of political interference with his company’s products. The concern is understandable: a government may purchase equipment, pay its subscriptions and fulfil its contractual obligations, only to discover that the operational value of its investment remains dependent upon decisions made outside its jurisdiction. European interest in alternative satellite communications providers, including Eutelsat’s network, reflects a wider recognition that dependence upon a single supplier creates strategic vulnerabilities. The problem is not peculiar to American technology, nor does it imply that European providers would be immune from political influence. Rather, it demonstrates why diversification and enforceable service arrangements have become essential considerations in defence procurement.

For Ukraine, the implications are particularly acute because satellite communications have become woven into the practical conduct of her national defence. Modern weapons systems are not merely physical objects whose effectiveness is determined at the moment of purchase. They depend upon software, communications networks, satellite navigation, maintenance arrangements, replacement components and continuing access to technical services. A tank may remain physically present upon the battlefield even when its supplier becomes politically hostile, although its long-term usefulness will suffer without spare parts and ammunition. A remotely controlled drone dependent upon a particular satellite network may lose much of its practical utility almost immediately if that network is restricted. The technological sophistication that gives modern armed forces their effectiveness can therefore create new forms of external dependency. Independence in military affairs increasingly requires not only possession of weapons, but meaningful control over the systems upon which those weapons depend.

This is one of the central dilemmas of contemporary international politics. The United States has long exercised extraordinary influence over the defence capabilities of her allies through the supply of sophisticated military equipment. In the past, that influence was exercised principally through governments, procurement agreements and military assistance programmes. Today private corporations occupy an increasingly important position within the same structure. Their executives possess technical powers that may have strategic consequences, while their governments retain political and contractual means of influencing corporate behaviour. The resulting distribution of authority is often opaque. It may be difficult even for a foreign government dependent upon a particular service to distinguish a commercial decision from a governmental instruction, particularly when both categories of authority operate through the same technical infrastructure.

The fundamental issue, then, is not whether Elon Musk should be regarded as uniquely responsible for Ukraine’s dependence upon Starlink. His company’s contribution to maintaining Ukrainian communications since February 2022 has been substantial, and Ukraine has repeatedly acknowledged its importance. Nor is the issue whether the United States is entitled to determine the conditions under which she supplies military technology to foreign governments. Every sovereign state necessarily retains a degree of discretion concerning its military assistance and international commitments. The more difficult question is whether restrictions upon essential wartime services should be introduced without warning, transparent authority or reasonable arrangements to prevent avoidable disruption. Even where a supplier has legitimate technical or security concerns, there is a substantial difference between negotiating operational limitations in advance and implementing them in circumstances where military personnel may discover the consequences only when their equipment ceases functioning.

What conclusions, therefore, can properly be drawn from the events of 14 October 2026? The evidence of restrictions upon Ukrainian Starshield operations is substantial and comes from several mutually reinforcing sources. The fact that particular military missions were disrupted also appears well established, although the full extent of the consequences remains uncertain. Ukrainian Defence Ministry sources attribute the introduction of the restrictions to SpaceX, but the available public record does not establish which individual or institution made the originating decision. Musk’s previous involvement in Starlink restrictions provides an obvious line of investigation, while his own statement concerning American governmental control of Starshield introduces an equally important reason to examine the role of the United States administration. The contemporary diplomatic dispute over Ukrainian attacks upon Russian energy infrastructure makes a political explanation plausible. It does not, by itself, make that explanation proven.

Indeed, the most revealing aspect of the affair may be that the identity of the ultimate decision-maker remains uncertain even after the practical consequences have become apparent. If Musk personally authorised the restrictions, the episode would raise questions about the strategic powers exercised by an unelected corporate executive. If the Trump administration instructed that the restrictions be imposed, it would raise questions about the use of military communications as an instrument of diplomatic pressure upon a country defending herself against invasion. If neither was directly involved, and the disruption resulted from technical or contractual decisions taken elsewhere, it would expose weaknesses in the arrangements governing the provision of critical wartime infrastructure. Each explanation leads to a different assessment of individual responsibility, but all three expose the importance of clearly established authority and dependable communications.

There is a final irony in the circumstances. Russia has spent years attempting to destroy Ukraine’s communications infrastructure through bombardment, cyber operations and electronic warfare. Ukraine, with considerable ingenuity and assistance from her international partners, has repeatedly found ways to overcome these difficulties. Yet the events of October 2026 suggest that some of her sophisticated military capabilities can also be impaired through decisions concerning access to technology developed thousands of miles from the battlefield. No Russian missile need strike a communications centre, and no Russian electronic warfare system need overwhelm a satellite signal. A modification to the conditions of service may be sufficient to frustrate an operation painstakingly prepared by Ukrainian soldiers.

The decisive question is therefore not merely who changed the settings of Starshield, but who possesses the authority to change them again, and upon whose instructions. Until that question is answered, Ukraine will remain exposed to an uncertainty that no military commander should welcome: the possibility that the effectiveness of essential equipment depends upon decisions made by persons whose identity, motives and accountability remain obscure. In earlier centuries, sovereign governments understood that control over strategic communications was inseparable from the exercise of military power. The war in Ukraine has demonstrated that this remains true in the age of artificial intelligence, autonomous weapons and satellite networks. The technology has advanced magnificently. The political problem of determining who controls it, and in whose interests, remains as old as war itself.