This essay is adapted from The Quantum Revolution: A Guide for Allied Policymakers, newly released by the Hoover Institution. A product of Hoover’s Allied Coordination Working Group and Applied History Working Group, the report offers a comprehensive guide to quantum computing, sensing, and communications, with special emphasis on why American leadership in the quantum era depends on orchestrating resilient networks of allied democracies. Download a copy here.
As speculation swirls about artificial intelligence, quantum technologies loom on the horizon as the next frontier. They promise potentially godlike capabilities: secure, instantaneous communications; sensors that can spot gold mines and submarines hiding on the seabed; algorithms that crack the world’s hardest cryptographic systems and lay bare its most sensitive secrets. Quantum stands a good chance of being the next world-shifting technology. But it is devilishly hard to understand. The subatomic world is a realm of spooky physics, intimidating jargon, and Greek symbols.
Is China winning this technological race? And what would it take for the world’s democracies to win? There is a temptation to cram quantum into the mental models we have built for AI. A two-horse sprint, with American and Chinese labs out front, capital and compute pooling behind the leaders. The analogy is tempting because the quantum computing race does partially resemble the early days of classical computing. But applying it too rigidly is a category error—and risks leading the United States to misallocate attention and money at the most dangerous possible time.
Our argument in this report is simple: the allied democracies can win the competition for quantum, but only if they work as a team and make sure all the bases are covered, while the key supply chains and enabling technologies behind quantum are still taking shape.
Off to the races
The first thing to understand is that there is no single quantum race. There are at least three, and they are running on different clocks:
- Quantum computing exploits the curious characteristics of atomic particles to attack problems that are intractable for classical machines.
- Quantum sensing uses the same physics to measure time, gravity, and magnetic fields with precision beyond anything classical instruments can reach.
- Quantum communication aims to move information near-instantaneously and in ways that make eavesdropping physically detectable.
These three fields share underlying science and a great deal of plumbing. But they are maturing at different rates, they have different supply chains, and they reward very different national strengths. According to various metrics, the United States currently leads in quantum computing. China is meaningfully ahead in quantum communications, having built the largest terrestrial key-distribution networks and demonstrated satellite links across thousands of kilometers. Sensing is a genuine three-way contest among the United States, Europe, and China. There is not likely to be one podium with a single gold medalist.
No one country is self-sufficient. The exotic hardware that quantum depends on—dilution refrigerators that chill processors colder than deep space, narrow-linewidth lasers, single-photon detectors, cryogenic control electronics—are produced by a thin, scattered set of suppliers concentrated in Europe and Japan. Finland makes some of the best refrigerators. The Netherlands and Switzerland make detectors and control stacks. Japan anchors the optics and lasers that nearly every modality needs. Many of these components are produced by single firms with specialized IP and process knowledge.
The supply chain is poorly mapped. China, having been slammed with export controls on semiconductors, is now racing to build self-sufficiency across the entire quantum supply chain—while positioning itself to make the Western democracies depend on Chinese hardware. This likely includes an all-out, state-sponsored campaign to access and copy Western IP, as well as huge investments to train the next generation of quantum talent in China. The allies have no equivalent strategy.
China does not have to win the entire supply chain to cause significant problems in quantum. It just needs to seize one chokehold.
Consider the example of chips. The Netherlands does not design or produce semiconductors. But a Dutch company, ASML, makes by far the best extreme ultraviolet lithography (EUV) machines. That single chokepoint gives Amsterdam leverage over an industry it could never have dominated outright. The quantum stack is full of latent ASMLs—potentially in cryogenics, in photon detection, error correction software, or control chips—that have not yet consolidated. Allied coordination is necessary now, before a dominant architecture emerges and locks the geography in place, shaping the chokepoints for the next decade and beyond.
Risks and uncertainty
The more vexing issue, which the marketing obscures, is that we don’t yet know what commercial tasks quantum computers will be good for. This is potentially a much bigger problem than industry admits, and it is another reason we need allied coordination.
Quantum computers will never be better than classical computers overall. They will be different: potentially unimaginably better at some tasks, but still far worse at others. There are two clear use cases: code-breaking (with the government the only plausible legitimate customer) and the simulation of quantum systems themselves. The latter approach could potentially be transformative for chemistry, materials, and drug discovery. Simulations on sufficiently advanced quantum computers could potentially design scarce materials from first principles, just as the discovery of the Haber-Bosch process freed Germany from its dependence on Chilean nitrates and let it fight on through a naval blockade in World War I. Beyond these two applications, though, it is far from clear that quantum computers could ever do valuable work faster and cheaper than classical computers.
Some of the best minds in theoretical computer science have spent thirty years trying to widen that landscape. They may yet succeed; algorithmic breakthroughs cannot be scheduled. But they cannot be assumed. Even if large-scale, fault-tolerant quantum computers arrive in the early 2030s, the companies that produce them may lack a commercial market to pay back their large capital investments. The quantum computing race is, at bottom, a geopolitical race. Even if the commercial applications never arrive, each side faces a national security imperative not to fall too far behind, in case the technology matures as the optimists hope. Allied governments should agree now to cooperate to ensure that allied quantum computing companies are in the lead, and in return should enjoy privileges of early access once they are deployable.
There is a second, compounding, source of uncertainty: quantum computing does not follow Moore’s law. For more than sixty years, classical computing has tracked a strikingly consistent trend along a single axis. Transistors keep getting smaller, faster, and cheaper, with the price roughly halving every two years for six decades, as if on a metronome. But quantum bits (“qubits”) are not the same as transistors, which simply carry 1s and 0s. Getting to usable (“fault tolerant”) quantum computers will require several things to improve at once: more physical qubits, longer coherence times for the qubits, higher gate fidelity, better error-correcting codes, and control systems fast enough to catch errors before they cascade. Progress in one dimension can mask stagnation in another. There is no one single benchmark to watch. They all bear on one another.
To be sure, the composite benchmarks have been improving at a startling clip. Quantum volume, which aims to provide a proxy for several other benchmarks, has been rising roughly 10x a year since 2018. That is far, far faster than Moore’s law. But extrapolating a clean exponential from a field with this many moving parts is a forecaster’s trap. Under aggressive assumptions, fault-tolerant quantum machines could appear in the early 2030s. But even a modest slowdown in any one input could push that out by many years. Anyone who quotes you the date with confidence is selling something, and the wild swings in quantum stock valuations are the market discovering this in real time. Investing in quantum computing companies is far riskier than investments in semiconductors in the 1950s and 60s. Feast could turn to famine, unless allied governments keep sending signals that maintain a favorable funding environment.
Allied coordination
If the timeline is so uncertain, why treat any of this as urgent? Because some of the risks are front-loaded onto today.
The clearest example is cryptography. A quantum computer large enough to crack standard internet encryption algorithms at scale does not exist. (It would have to be at least 100,000 qubits, according to one recent estimate, but potentially much larger. The honest answer is that we don’t know.) Still, encrypted data can be stolen now and stored, then decrypted whenever a capable computer comes online. “Harvest now, decrypt later,” they call it, potentially exposing secrets with a long shelf life. This is why, to hedge, allied governments must accelerate the shift onto post-quantum cryptography: new classical algorithms, which are already standardized, that can be slotted into existing networks and make them more resilient to future quantum attack. The federal migration target is 2035, at an estimated cost of $7 billion. Allied governments should be moving faster—and together.
One useful lesson from semiconductors is that early advantages in certain nodes of the supply chain tend to compound as demand scales up. In quantum sensing, for example, the first movers to field jam- and spoof-resistant navigation—positioning that doesn’t depend on satellites an adversary can blind—will build wide moats. Standard-setting will be critical. The choices being made right now about what programming languages, interfaces, and access models to use will harden into norms long before any consumer market exists, just as NVIDIA’s CUDA API quietly locked developers into its ecosystem. For classical computing, standards were set by scientists and other technocrats, without much regard for geopolitics. But this was because the Soviet Union wasn’t really a player. Today, China will do everything it can to bend global quantum standards in ways that advance its own companies, just as it worked effectively to write rules of 5G that favored Huawei. The allies need to work together to ensure the rules are fair.
The bottom line is that the United States is unlikely to dominate quantum the way it now dominates AI. The field is too distributed, the supply chain too dependent on allies, the decisive applications too speculative, and the diffusion of any breakthrough too fast to bottle up. A small, internationally networked research community will not keep secrets for long; a breakthrough at Stanford or Cambridge will be global knowledge within months.
The task, then, is not to win a race but to orchestrate an alliance—and an alliance implies a division of labor. Its rough shape is already latent in the map. The United States has the platform, software, and integration layers. It is best placed to bankroll the hunt for the missing algorithms. Europe and Japan anchor the hardware chokepoints—refrigerators, lasers, optics, single-photon detectors—on which every other player depends. Britain, the EU, and Australia are leading in much of the sensing and GPS-free navigation work.
Coordination means hardening each lane on purpose, rather than leaving it to the accident of where a startup happened to incorporate—and without asking allies to surrender the sectors they lead. It means treating the allied supply chain as a shared strategic asset rather than a set of foreign vendors to be managed warily. It means coordinating export controls so they protect genuine chokepoints without strangling the collaboration that allied progress depends on. It means fixing the self-inflicted wounds—visa and immigration rules that drive away the foreign-born talent on which the American quantum ecosystem disproportionately runs, and that could just as easily seed the next generation of startups in Singapore or Zurich. And it means funding the basic research, at universities and national labs, where the missing algorithms will come from if they come from anywhere.
The race framing flatters American strengths and hides American dependencies. The real work of building a resilient coalition is less satisfying than the prospect of a decisive win. But it is the only viable strategy that fits the technology we actually have.
Adapted from Eyck Freymann, Sebastian Orbell, Sophie Coste, and Katharina Klotz, The Quantum Revolution: A Guide for Allied Policymakers (Hoover Institution, 2026).
