In 1982, Toshiba Machine Company and Norway’s Kongsberg Vaapenfabrikk sold nine precision milling machines to the Soviet Union for roughly seventeen million dollars. The machines were disguised in export documents as simple equipment. What they actually did was allow the Soviets to carve submarine propeller blades to tolerances measured in microns. American submariners had a nickname for Soviet submarines: “roaring cows,” because they could be detected from a hundred miles away. After those nine machines, Soviet submarines became twenty times quieter. Detection range collapsed from a hundred miles to about five. The US Department of Defense estimated it would cost thirty billion dollars to compensate. Seventeen million dollars in, thirty billion out - a strategic disadvantage ratio of nearly two thousand to one.
This is the central paradox at the heart of technology denial regimes - the subject of my recent paper, “A Strategic Assessment of Technology Denial Regimes” (Takshashila Discussion Document 2026-04), for the Takshashila Institution. Countries use export controls and supply chain restrictions to gain strategic advantage over rivals. And yet, the history of these attempts is littered with unintended consequences so dramatic they border on self-parody. Understanding why requires looking not just at when these regimes work, but at the strange ways they tend to fail.
The Things That Actually Matter
My research across decades of technology denial - in nuclear, space, computing, and defence sectors - suggests that effectiveness comes down to two variables: how much of the supply chain the imposing country controls, and how much multilateral cooperation it can sustain. When both are high, you get what I call the “Iron Gate” - the kind of near-impenetrable barrier the Nuclear Suppliers Group built around uranium enrichment after India’s 1974 nuclear test. When both are low, you get the “Hollow Threat” - like the 1990s attempt to restrict strong encryption, which collapsed because, as I put it in the paper, cryptographic schemes are just mathematics, and any competent programmer can implement them. No export control can stop an idea.
The problem is that most real-world cases fall in between, and the regimes that start as Iron Gates tend to erode into something much weaker. CoCom - the Cold War-era export control bloc that ran from 1949 to 1994 - is the most instructive example. It was never a treaty. It was a gentlemen’s agreement, with no legal force and no enforcement mechanism, run out of a room in the US Embassy annex in Paris. Most people didn’t even know it existed. When France sold ten machines similar to the Toshiba ones to the Soviets and faced minimal consequences, the moral hazard was never addressed. If one member could defect and pay no price, why would anyone else hold the line?
For targeted countries, what matters, as the other framework in my research paper identifies, is indigenous capabilities of the country, which could be research and development (R&D) capabilities or manufacturing capabilities or any other which provides it ways to counter denial regimes. What is also important is the ability to form alternative partnerships which can help it overcome the impact of these controls, as we will see later in this essay. And all of this will fail unless feedback loops are incorporated into the entire process. These feedback loops will assess a targeted country’s abilities and state of development, to recalibrate controls. This recalibration could also see the forthcoming of certain incentives to drive the expected behaviour from targeted countries. On the part of the targeted country, this means it should assess its own state periodically to understand whether it can achieve some amount of technological sovereignty which can help it overcome tech denial regimes. If it is not up to the mark, then there have to be strategic pivots towards more partnerships or negotiations.
The Boomerang Effect
There is a pattern that repeats across technological eras with almost eerie consistency, and I call it strategic reversal. In 1950, the United States imposed a near-total embargo on China after it entered the Korean War. The logic was straightforward: China is technologically backward, isolate it and contain it. What actually happened is that the embargo pushed China directly into the Soviet Union’s arms. The Soviets responded with 156 major industrial projects - factories, blueprints, trained engineers, entire knowledge systems transferred wholesale. When the Sino-Soviet split came in 1960 and Soviet support evaporated, China had enough of a foundation to go it alone. By 1964, it had nuclear weapons. By 1970, it had put a satellite in orbit. The United States had, through the act of trying to contain China, inadvertently funded and accelerated the creation of the strategic competitor it feared most.
India offers a more optimistic version of the same story. When the Missile Technology Control Regime cut ISRO off from key propulsion and launch technologies, India didn’t simply try to replicate what was denied. It innovated around the constraint, eventually developing low-cost satellite launch capabilities that now compete globally and undercut established providers on price. The constraint became the innovation driver. What began as vulnerability became competitive advantage. India’s space trajectory is probably the most hopeful illustration of what my paper calls the “Resilient Transformer” - a country that combines strong indigenous R&D with selective alternative partnerships to emerge from denial stronger than before.
The War Beyond the Customs Desk
The most consequential technology battles today are not being fought at the export licensing desk. They are being fought through financial infrastructure - when Iran was disconnected from SWIFT in 2012, it lost the bulk of its oil revenue almost immediately, not because any technology was denied but because the payment rails connecting it to global commerce were switched off. They are being fought through standards bodies - China’s early and aggressive entry into 3GPP, which sets 5G telecommunications standards, means that technical requirements beneficial to its ecosystem are now embedded into the global baseline. And they are being fought through talent - knowledge moves with people, and the country that attracts the right researchers imports capabilities that no export control list can screen.
None of this means technology denial is pointless. The nuclear regime has been genuinely effective at slowing proliferation for decades. Semiconductor controls have real bite in the short term. But the paper’s conclusion is sober: the primary long-term effect of most technology denial regimes is not stopping rivals from advancing, but accelerating the fragmentation of global technology markets and catalysing the very alternative innovation ecosystems that ultimately erode the imposing country’s advantage. The boomerang comes back. The question is only how long it takes, and whether the country that threw it is ready when it does.