The Soviet Unions Secret Internet That Almost Changed History
by Scott
In October 1970, a Soviet mathematician named Viktor Glushkov walked into Stalin’s former office in the Kremlin to present what he believed was the most important proposal of his life. He was there to ask the Politburo to authorize and fund the construction of a nationwide computer network that would connect tens of thousands of computing centers across the Soviet Union, linking factories, farms, government ministries, and economic planners into a single real-time information system capable of managing the largest planned economy in the world. The system he envisioned would anticipate cloud computing by decades, would process economic transactions electronically in ways that prefigured digital finance, and would give Soviet economic planners access to information about the state of the economy that was, at that moment, arriving weeks or months late through a paper-based bureaucracy that Glushkov believed was slowly strangling the country.
He had been working toward this moment for nearly a decade. His proposal had the support, or so he believed, of General Secretary Leonid Brezhnev and Premier Alexei Kosygin, the two most powerful men in the Soviet state. He entered the room and noticed that both of their chairs were empty. Their absence that day, explained away as prior commitments to other state functions, emboldened the opponents of the plan in the room to move decisively against it. Finance Minister Vasily Garbuzov, who understood that the network would transfer economic information and therefore economic power away from his ministry and others like it, pushed through a counterproposal that gutted the plan. Permission was granted to install computers at individual production centers, but the critical element, the networking of those centers into a unified system, was refused. The Soviet Union would not build its internet. The reasons why are a story that reveals as much about institutional power and bureaucratic self-interest as they do about technology.
To understand the significance of what was rejected in that room, it is necessary to understand the problem it was designed to solve. The Soviet command economy operated through a system of central planning that required the collection, processing, and distribution of an almost incomprehensible volume of economic information. Every factory in the country produced outputs and required inputs that had to be coordinated with thousands of other enterprises, and the coordination was managed through plans developed by a central planning authority called Gosplan. By the late 1950s, the complexity of this coordination problem had grown to a scale that was beginning to defeat the bureaucracy charged with managing it. Glushkov calculated that if the paper-driven methods of economic planning continued unchanged, the planning bureaucracy would need to expand by nearly forty times its existing size by 1980 just to keep pace with the growing complexity of the economy. The arithmetic of the problem was becoming impossible.
The idea of using computers to address this problem did not originate with Glushkov. It originated with a military engineer named Anatoly Kitov, a colonel in the Red Army who had worked on some of the earliest Soviet computers in the 1950s and who had arrived, through practical experience with the limits of human information processing, at the conviction that only networked computing could solve the coordination problems of the planned economy. In 1959, Kitov sent a letter directly to Nikita Khrushchev proposing the creation of a unified national network of computer centers. The proposal, which became known as the Red Book project, was practical in its ingenuity. Kitov recognized that the military had already invested heavily in computing infrastructure for defense purposes. He proposed that this infrastructure could serve a dual purpose: processing military information during wartime and peacetime defense operations, while during the hours when military demands were low, the same computing capacity could be used to process economic planning data for civilian purposes.
The proposal was intercepted by military authorities before it reached Khrushchev. The reaction was fury. The thought that military computing resources should be shared with civilian economic planners was experienced not as a practical suggestion for maximizing the utility of existing infrastructure but as an attack on military institutional autonomy and an implicit criticism of the army’s management of its resources. Kitov had also, in the same letter, criticized the efficiency of both military and civilian administration, which compounded the offense. A secret military tribunal was convened to review what was characterized as a breach of trust, and its verdict was swift: Kitov was stripped of his Communist Party membership and permanently dismissed from the military. The first serious proposal for a Soviet national computer network ended in the professional destruction of its author.
The idea survived its author’s punishment, carried forward by Glushkov, who became director of the Institute of Cybernetics in Kiev in 1962 and who transformed what had been Kitov’s practical proposal into something considerably more ambitious. Glushkov’s vision for the All-State Automated System, which the Russians abbreviated as OGAS, was not merely a management tool but a comprehensive reimagining of how a socialist economy could function. The system he proposed would have consisted of a three-tier hierarchy: a central processing hub in Moscow at the top, up to two hundred regional computing centers in major cities at the middle level, and up to twenty thousand local terminals at economically significant points throughout the country at the base. These centers would be connected through existing telephone lines and new dedicated links, allowing real-time information to flow up and down the hierarchy and, crucially, horizontally between any authorized users anywhere in the network.
The technical architecture of the OGAS was in several respects ahead of its time. Glushkov understood the value of decentralized design, building a system in which any authorized user could contact any other user without requiring explicit permission from the central node for each transaction. This was not merely an engineering convenience but a philosophical commitment to leveraging local knowledge and local expertise in the operation of the network, rather than funneling everything through a bottleneck of central authorization. The system also contemplated what would today be called electronic money, a proposal for virtualizing currency into an online ledger of accounts that would make physical cash unnecessary for transactions within the planned economy. Glushkov described this in the early 1960s as a faithful fulfillment of Marxist theory about the eventual supersession of money in a communist society, framing an apparently radical technical proposal within ideologically acceptable terms.
The estimated cost of the OGAS was enormous, requiring tens of billions of rubles and decades of implementation work. Glushkov believed the return on this investment would be several times its cost over the first fifteen years of operation, but the upfront numbers were staggering by any measure. The American government was paying close attention. In 1962, the US government assessed the OGAS project as a major threat, and Arthur Schlesinger Jr, then serving as a special assistant to President Kennedy, described a full Soviet commitment to cybernetics as potentially giving the Soviet Union a tremendous advantage in economic productivity and industrial capacity. This assessment, coming from the heart of the Kennedy administration, suggests that American observers took the Soviet network project more seriously than the Soviet bureaucracy ultimately would.
Glushkov spent most of the 1960s building political support for the OGAS, traveling between Kiev and Moscow, visiting more than a hundred industrial sites to understand how the network would need to function in practice, and cultivating allies within the government who could shepherd the proposal through the labyrinthine Soviet bureaucratic system. He was genuinely skilled at this, more politically adept than Kitov had been, and he accumulated considerable prestige in the process, receiving major Soviet honors and being elected to the Academy of Sciences. His institute in Kiev attracted young researchers with an average age of around twenty-five, and the energy and ambition of the place was remarkable. The staff even created a playful imaginary entity they called Cybertonia, a fictional country run by a council of robots, complete with its own constitution and currency, a speculative design project that imagined the cybernetic future they were working toward.
By 1970 the OGAS proposal had been refined through years of development and was ready for the decisive moment. Glushkov had reason to believe that Brezhnev and Kosygin, both of whom had expressed support for the project in private, would provide the political cover needed to push it through despite the opposition from various ministries. The empty chairs at the October meeting signaled that this support would not be publicly expressed when it was most needed. Garbuzov’s counterattack succeeded in stripping the proposal of its essential networking component, reducing it to something that could not threaten the ministries’ control over their own information and resources.
The bureaucratic logic of the opposition deserves careful examination, because it was not irrational from the perspective of those who exercised it. The OGAS threatened something that every ministry in the Soviet system valued above almost anything else: control over information. In a planned economy where power derived from knowledge of economic conditions, resources, and bottlenecks, a system that made that information universally accessible within the network would have fundamentally redistributed the power that came from controlling it. The Ministry of Finance knew exactly what the OGAS would mean for its influence: a reduction in the informational asymmetry that was one of its primary sources of leverage. Garbuzov’s opposition was not merely obstruction for its own sake. It was rational self-preservation by an institution that understood precisely what the proposed reform would cost it.
This dynamic has been analyzed extensively by media scholar Benjamin Peters in his 2016 book on the Soviet internet, and his central observation has a sardonic precision that captures the essential irony of the situation. The Soviet Union failed to build its internet not because the institutions of state socialism could not cooperate, but because those institutions, when faced with a technology that would require genuine cooperation and genuine sharing of information, behaved like competing capitalist enterprises protecting their market positions. Meanwhile, in the United States, ARPANET, the network that would eventually become the internet, was being built through exactly the kind of centralized state funding and collaborative institutional behavior that the Soviet system was nominally organized to facilitate. The Americans built their network through something resembling socialist coordination. The Soviets failed to build theirs because their nominally socialist institutions would not cooperate.
Glushkov did not give up after the 1970 meeting. He continued to advocate for scaled-down versions of the OGAS concept through the 1970s, pursuing successor proposals that faced the same institutional opposition and received the same inadequate funding. The years of fruitless advocacy took a toll. A man who had been rigorously rational and analytically precise began, toward the end of his life, to suggest that American spies had sabotaged his work, and that a turbulent emergency landing on a flight he took shortly after the 1970 Politburo meeting had been an assassination attempt against him. Whether these suspicions reflected genuine evidence or the strain of prolonged institutional defeat is impossible to determine at this distance. Glushkov died in 1982, having spent two decades on a project that never came close to realization, and having published a book about the OGAS in the year before his death intended for Soviet schoolchildren, an audience that might do what his own generation had refused to.
The timing of the project’s failure matters as much as the fact of it. Had the OGAS been approved in 1970 and begun implementation, the Soviet economy in the mid to late 1970s would have been operating with an information infrastructure that might genuinely have altered the trajectories of Soviet economic difficulties. Glushkov himself had warned the Politburo in 1970 that if the full OGAS was not implemented, the Soviet economy would encounter severe difficulties in the second half of the 1970s and the leadership would be forced to return to the question of network-based economic management anyway. The difficulties arrived more or less on the schedule he predicted. The leadership did not return to the question in any serious way. The flow of oil revenue following the 1973 price shock flooded the Soviet budget with hard currency that temporarily masked the structural problems Glushkov had identified, providing bureaucrats with a decade of reduced urgency that further delayed any fundamental reform of economic management.
By the time Soviet leadership under Gorbachev began seriously attempting the reforms that might have included something like what Glushkov envisioned, the institutional and technical foundations for such a project had been largely abandoned, the expertise had dispersed, and the political window for transformation was closing in ways that would soon make the entire question irrelevant. The Soviet Union obtained its first connection to the international internet in 1990, through a private telecommunications enterprise called Relcom that established a telephone connection to Finland. It was a modest, commercial, externally dependent connection, almost the precise opposite of the ambitious, internally developed, nationally sovereign network infrastructure that Glushkov had spent his career trying to build.
The OGAS is the most dramatic example of a broader pattern that characterized Soviet engagement with information technology: a genuine capacity for visionary thinking about the potential of computing, combined with an institutional structure that systematically prevented the realization of that vision. Soviet scientists and engineers were, in the 1950s and 1960s, genuinely at the frontier of thinking about what networked computing could accomplish. The ideas were often original, sometimes prescient, and occasionally ahead of their Western counterparts. The problem was never primarily technological. Soviet computing lagged behind American computing in hardware and in some areas of software, but this gap was not what killed the OGAS. What killed it was an institutional environment in which the sharing of information was experienced as a threat by the entities that controlled information, and in which the coordination required to build a genuinely nationwide system could not be achieved because the incentives of every major actor pointed toward obstruction rather than cooperation.
The story of the Soviet internet is a reminder that major technological systems do not come into being because they are technically possible or even economically rational. They come into being because the institutional arrangements and incentive structures of the societies that might build them allow the necessary cooperation to occur. ARPANET worked in part because the Defense Advanced Research Projects Agency could fund it centrally and because American universities had a culture of collaborative research that made sharing computing resources seem natural rather than threatening. OGAS failed in part because the Soviet ministry system had developed a culture in which information was power, sharing was loss, and any reform that threatened institutional positions would be fought with the full force of institutional self-preservation.
What the world might have looked like had the OGAS been built and operated successfully is genuinely unknowable, but the question is not entirely idle. A Soviet economy with real-time information feedback, capable of identifying and correcting inefficiencies continuously rather than through a planning process that was always working with months-old data, might have been considerably more competitive than the economy that actually existed. Whether such a system could have addressed the deeper incentive problems of central planning, the impossibility of replacing the distributed price signals of a market with any centralized information system, is a question that economists continue to debate. Glushkov believed it could. The question was never tested. The empty chairs decided that.