Quantum Unfiltered #10 — America Just Made the Case for Quantum Sovereignty
Washington switched off a US company's two best AI models for the entire world in an evening. The same machinery applies to quantum, and the dependency runs deeper.
In this edition: On Friday evening, the US government ordered Anthropic to shut off its two most capable AI models for every customer worldwide, within hours, no warning. I explain why that is the cleanest argument yet for quantum sovereignty, and why buying a US machine outright does not escape the problem. A short detour into helium-3, the coolant a whole class of quantum computers can’t run without and the subject of much discussion in recent weeks, shows how deep the dependency reaches. I point to my two new build deep dives that take you from an empty lab to a working machine, the older sovereignty deep dive that mapped this terrain, and the QuantWare raise that shows the componentized supply chain forming in real time. My book Quantum Sovereignty ships in a few days, with Quantum Systems Integration as its engineering companion.
America Just Made the Case for Quantum Sovereignty
At 5:21 p.m. Eastern on Friday, June 12, Anthropic received a letter from the US Department of Commerce. By that evening, two of its most capable AI models were dark for every customer on the planet. Not throttled, not geofenced. Off. Anthropic did not choose to pull them; its own government ordered it to, through an export-control directive barring any foreign national, anywhere, from using the two models. Since no provider can separate foreign nationals from everyone else in real time, the only way to comply was to switch them off for the entire customer base, Americans included.
I set the AI-policy fight aside in my full analysis, because for the argument I care about it doesn’t matter whether the order was justified. For the past year, American quantum founders, investors, and officials have asked me a version of the same question: why are so many of your clients asking about independence from US quantum technology? It is the best in the world. What is the worry? I never had an example clean enough to make the answer land. On Friday evening, Washington provided one.
The shape of the action is what matters, and it holds whether the underlying call was sound or mistaken. Speed: the directive arrived at dinner and the models were gone the same night, no grace period to move workloads. Reach: it turned not on where you sit but on who you are, sweeping up a German pharmaceutical firm, a Brazilian bank, a Japanese manufacturer, whatever data center they use. Recourse: the vendor objected plainly, called it a misunderstanding, and complied within hours anyway, because a provider’s obligations to its own government sit above its contracts with its customers.
Here is why this lands harder for quantum than for AI. The same export-control machinery applies, and the dependency underneath is heavier. The industries that will need quantum compute most urgently, the narrow set where quantum delivers real advantage, pharmaceuticals, chemicals, batteries, advanced materials, are also the ones most central to national security. I called this the utility trap in my Quantum Utility Map. And buying the hardware outright does not escape it. A quantum computer bought today from a full-stack vendor is, in most respects, still that vendor’s machine: the processor was fabricated abroad, the control electronics and cryogenics came from abroad, and calibration and firmware updates are performed by the vendor’s own engineers under the service contract, which makes the firmware-update policy a form of remote control. The buyer had a quantum computer. It did not have control of one.
The full piece covers the two-ways-to-read-it framing, why Paris couldn’t stop talking about sovereignty at the Q-Day Summit last week, and what I tell the Americans in the room. Read it on PostQuantum.com. The short version: “trust us, we will not turn it off” stopped being a sufficient answer the moment the world watched the United States turn it off, on its own flagship company, over a finding the vendor calls minor.
For a year I have argued this case to American founders and officials without an example clean enough to land it, and I keep a careful distance from the quantum panic industry that inflates exactly this kind of fear. So let me be precise about the claim. I am not saying a quantum computer will break your encryption next year, or that every firm needs quantum access today, or that American technology is a trap. The claim is narrow: the dependency is coming for specific industries on a knowable path, and the mechanism for revoking it already exists and was just used in public. You only need to believe that some industries will come to need quantum compute, and that the supplier’s government can switch it off. Both halves of that sentence are now demonstrated.
A Note From Below the Qubit: The Helium-3 Question
Before Friday, the sovereignty conversation I kept having in recent weeks was about something far less abstract than export law: a gas. Helium-3 has been a live topic in quantum circles since Bluefors signed a $300 million contract to buy it mined from the Moon, and it is worth a short detour here because it shows how deep the dependency Friday exposed actually reaches, down to a single isotope that a large class of quantum computers cannot run without.
The physics is unforgiving. Superconducting and silicon spin qubits operate near 10 millikelvin, and only a dilution refrigerator gets them there, a process that depends on helium-3 specifically, since ordinary helium-4 cannot do the job. Helium-3 is barely present in nature; almost every gram is a decay product of tritium recovered from nuclear weapons stockpiles, which puts the supply in the hands of a few governments (the US, Canada, and Russia until 2022) and leaves global production running well short of demand. That is a sovereignty exposure two layers below the processor. The scarcity is real, which is the half of the story worth taking seriously. The other half, the projections that justify mining the Moon, rests on two errors I take apart in the full explainer: a dilution refrigerator recirculates its helium-3 in a sealed loop rather than burning it, so demand is a one-time charge per machine, not an annual rate, and only some modalities need it at all (trapped-ion, neutral-atom, and photonic systems largely don’t). The real pressure on helium-3 comes not from the Moon but from magnetic cooling, which is advancing fast. The reason it belongs in this edition: when even the coolant in your fridge traces back to a handful of national nuclear programs, architectural control means accounting for layers most buyers never think to ask about.
From an Empty Lab to a Working Machine
If Friday made the case for architectural control, the question becomes how you actually achieve it. That is the subject of the two build deep dives I recently published on PostQuantum.com.
How to Build a Quantum Computer walks the full journey from empty lab to first qubit signal across every major modality, with companion articles on the parts that determine whether you control the machine or merely host it: cryogenic infrastructure and the He-3 supply, control systems, the quantum OS and orchestration layer, HPC integration, and supply chain concentration risk. The companion piece, What It Takes to Build a Quantum Computer, covers the strategic and economic side: what each layer costs, where the chokepoints sit, and what the build-versus-buy decision actually involves.
Read together, they make a sovereignty argument by construction. A nation does not need to fabricate every layer itself; what it needs is architectural freedom. The stack can be sourced in layers, a QPU from one supplier, cryogenics from another, control electronics from a third, then integrated in-house or by a neutral systems integrator. This is the same shift from sealed mainframe to open, modular machine that handed sovereignty from the producer to the user in classical computing, and I laid out the case for it in my older Quantum Sovereignty and Self-Reliance deep dive.
The Componentized Supply Chain Is Already Forming
The good news for any sovereignty-minded buyer is that the modular alternative is not theoretical. It is being built right now, and the clearest signal is where the money is going.
QuantWare, a Dutch QPU manufacturer, closed a $178 million Series B backed by Intel Capital and In-Q-Tel. Its entire business model is selling quantum processors as off-the-shelf products to customers who integrate them into their own systems with their own cryostats and control electronics. That is the open-architecture model in commercial form, and as I argued back in edition #8, the $14 billion now flowing into quantum is building a QPU component supply chain rather than monolithic machines. A buyer who wants architectural control no longer has to build every layer from a national-lab cleanroom. The components are becoming products, the integrators are emerging, and the multi-vendor machine is becoming a procurement option rather than a research project. Friday is the argument for choosing it.
Two Books, Out This Week
I’ve spent a long stretch of my working life on this dynamic, how critical industries come to depend on concentrated quantum supply chains and what it costs to climb back out. It is the subject of my book Quantum Sovereignty, out in a few days. I did not expect the cleanest illustration of the thesis to arrive from Washington the week before it published.
Its engineering companion is Quantum Systems Integration, the practical guide to the open-architecture alternative this edition describes: how to source each layer of the quantum stack, how to manage the multi-vendor integration, and what it costs at three different tiers. If Quantum Sovereignty makes the case for why architectural control matters, Quantum Systems Integration is how you build it. Sign up at both sites to be notified at launch.
A disclosure: my company, Applied Quantum, is a quantum systems integrator. We assemble quantum computers from modular components for clients who want architectural control, and most come to us because of sovereignty concerns. On Saturday morning, less than eighteen hours after the directive, a prospective client called to ask for a proposal. Friday did not create the demand. It removed the last hesitation. I have a commercial interest in this argument and I am being direct about it, because the analysis should be evaluated on its evidence, not on whether the person making it also builds the alternative.
The next time someone asks me why the rest of the world wants its own quantum stack, I won’t reach for a hypothetical about a distant Q-Day. I’ll point to a Friday evening in June, a letter that took effect by dinner, and two of the most advanced AI models on the planet going dark for everyone at once because a government decided they should. The quantum version of that letter has not been written yet. The countries and companies building their own machines are making sure that, when it is, they are not the ones waiting by the inbox.
If you found this edition useful, forward it to someone weighing a quantum procurement decision. If I got something wrong, hit reply. I read everything and correct publicly.
— Marin