We're not doing a CBDC.
That sentence is true. It is also the most load-bearing piece of misdirection in money right now.
The front door — a government digital currency in your pocket — got bolted shut, loudly. The back door — privately operated, state-regulated, programmable rails — is being built in public, quietly. The same capabilities arrive either way.
Deny the retail token. Build the rail underneath. Money that is traceable, conditional, and steerable at scale arrives without anyone ever issuing "a CBDC" — and the denial stays technically true the entire time.
There are two CBDCs. Conflating them is where the fog comes from.
Almost every headline argument runs on a confusion between two completely different things that share three letters.
A retail CBDC is a digital dollar in your pocket, issued straight by the central bank. Across the G7 it is stalled or dead. The United States didn't just decline it — it prohibited it. Canada wound its program down. The UK's "digital pound" is stuck in design; the digital euro is perpetually "in preparation." This is the thing being loudly denied. This is the decoy.
A wholesale CBDC is central-bank money that banks use to settle with each other. It is alive, funded, advancing, and almost never mentioned in the public fight. The denial is precise about the first kind and silent about the second. That gap is the whole game.
Read that twice. The program isn't cancelled, it's shelved with the parts kept oiled. The option is preserved on purpose. The "no" is a "not yet," stated out loud, and almost nobody heard the second half.
"Private rails" is not a metaphor. The buildout has names.
While the retail token gets denied, a parallel system of real, regulated, named infrastructure is being assembled. These are not predictions. They exist now.
The harmonisation is real. But it's a press release, not a back room.
The instinct is to imagine a secret meeting. There is no secret meeting. There is a project with a logo.
Project Agorá, convened by the Bank for International Settlements, brings seven central banks — the Bank of France for the Eurosystem, the Bank of Japan, Bank of Korea, Bank of Mexico, Swiss National Bank, Bank of England, and the New York Fed — onto one platform, with the Bank of Canada joining and more than 40 financial institutions including JPMorgan, HSBC, Deutsche Bank, Swift, Mastercard and UBS.
Its stated aim, in the open: integrate tokenised commercial-bank deposits with tokenised central-bank money — wholesale CBDC — in a single "network of networks." As of May 2026 it is advancing to real-value testing. It is published.
This matters: coordination is not being hidden, it's being filed under a different word. Not "control," but "interoperability." Not "a CBDC," but "cross-border efficiency." You don't need a conspiracy when the policy is announced and simply read as plumbing.
Trace one payment and you find four control surfaces. Each is owned by someone.
"Programmable money" makes people picture coins that expire. That's the smallest part. Follow a single token through the rail and the leverage is everywhere:
The famous fear fixates on surface 3. But the real grip is at 2 and 4: you never need the money itself to be programmable if you control who can hold it and who can cash out. That is already how sanctions and AML work. Tokenisation doesn't invent the capability. It changes the resolution and the speed — from slow, human, and exceptional to instant, automatic, and granular.
Efficiency and control are not two features. They are the same feature, pointed two ways.
The exact smart-contract capability that delivers the benefits is the capability that delivers the steering. There is no version that has one without the other, because they are the same line of code.
- Instant settlement, no waiting days
- Atomic delivery-versus-payment
- Automated, cheaper compliance
- Cross-border in seconds
- Money that expires on a date
- Spend restricted to approved categories
- Geofenced or per-wallet frozen
- Conditions enforced before it moves
Hold onto the one distinction the officials hide behind: programmable payments (conditions on a transaction) versus programmable money (conditions baked into the unit itself). They publicly disown the second. The first — already being built — gets you most of the way there.
"Private" isn't the cover story. It's the only design that doesn't blow up the banks.
If the goal were control, why not just issue a retail CBDC directly? Because a direct retail CBDC guts the commercial banking system.
If everyone can hold risk-free central-bank money, deposits flee the banks at the first wobble — an instant, frictionless digital bank run. Banks lose the deposit base they lend against. Central bankers know this and are genuinely afraid of it.
So the architecture being built isn't a choice of control over efficiency. It's the resolution of a structural conflict: keep the banks alive (they hold the deposits and mint the tokens) while still getting unified, programmable, traceable rails (the standards layer, with wholesale CBDC underneath). Which means you don't need to assume deception to predict the exact outcome. The incentives alone build it. That is a stronger claim than a plot, not a weaker one.
"We're not doing a CBDC" is load-bearing. It does three jobs at once.
It demobilises the opposition
Privacy and liberty energy organises against retail CBDC — the one thing that isn't happening — while the wholesale and tokenisation buildout sails past unwatched. The fight is aimed at the decoy by design.
It keeps every official statement true
Nobody has to lie. Each denial is narrowly accurate, so fact-checks "confirm" it, and anyone describing the larger shape gets filed under conspiracy. Technical truth as a shield.
It clears the runway
Because the rails are neutral substrate, a retail CBDC becomes plug-and-play later — when a crisis, a war, or a "we need to move stimulus fast" makes it convenient. The denial today doesn't foreclose the issuance tomorrow. It builds the road and leaves it empty. The present-tense honesty is exactly what makes the future-tense capability frictionless.
Infrastructure capture, then weaponisation, has already run twice.
You're not predicting a novel evil. You're predicting that a pattern which has happened before will happen again, with sharper tools. Each time, the rail was sold as neutral plumbing; each time, the power to deny came later and was never the thing debated when the pipes were laid.
The move isn't only about money. One of these shipped in the machine you're reading this on.
Everything above is payment rails. But "sold as neutral, the power to deny arrives later" is not a fact about money. It's a fact about who owns the pipe. Here is the identical shape, running right now, one layer down, in the operating system.
Every Windows installation is stamped, at setup, with a permanent Global Device Identifier, a GDID. Microsoft describes it the way SWIFT described messaging: neutral plumbing. Diagnostics. Crash reports. Feature-usage. Catching one machine that claims a free trial a thousand times. All true, all documented, none of it hidden. It is the machine's signature, and it does not change when you change your network, your country, or your VPN.
Then the power to deny, here the power to identify, arrives later, through an ordinary request. In July 2026 prosecutors unsealed how the FBI found an alleged member of the Scattered Spider extortion crew. He had done the tradecraft: a VPN, tunnelling tools, addresses rotated across countries. The VPN hid his IP. It could not touch the GDID. On a court order, after a criminal referral from Microsoft, the identifier was resolved to his own Snapchat, Apple and Facebook logins in Tallinn, New York and Thailand. Different networks. Same machine. Same person.
Read it as another row in the same table, off the money rails entirely. That is the point: the pattern needs no money, no plot, and no villain. It needs a documented capability that ships as maintenance, and a lawful door that opens onto it afterwards. Microsoft answered a warrant, and that is not the scandal. The scandal is that a permanent, state-reachable identity sits on every machine, governed in public by almost nothing: no disclosure policy specific to it, no user opt-out, no transparency line that counts the requests.
The GDID and Microsoft's own description of it. The court order, the criminal referral, the VPN that didn't mask it, the cross-platform login match. The absence, in public, of any policy governing when it is disclosed.
"Microsoft is an arm of the FBI." "It's a secret backdoor." "They handed over his entire digital life."
You need none of it. A documented feature and a lawful request produce the identical outcome, with no planted backdoor and no villain. And the accused may well deserve catching, which is exactly why the capability is easy to wave through.
Filed in full, with the complete firewall, as The Diagnostic.
Name the losers, in order of exposure.
- Anyone whose legal activity is unpopularThe Visa/Mastercard deplatforming logic — now instant, automatic, granular. Sex work, gambling, gun sales, dissident orgs, protest funds. Canada already did this by hand in 2022.
- The unbanked and cash-dependentAs cash acceptance erodes, the one exit that doesn't phone home disappears for the people who needed it most.
- Small banks and independent fintechsToll-payers on someone else's pipe, or gone. Compliance-by-scale eliminates them automatically — no hit list required.
- The privacy-preferring majorityNot by a dramatic ban, but by the slow disappearance of any default-private way to transact at all.
Where the airtight argument stops and the overreach begins.
This is the line that decides whether the whole thing is bulletproof or dismissible. Everything on the left is documentable. The moment you cross to the right, you've handed every critic the one thread that unravels it — and traded an argument you can't refute for one you can.
The projects, the laws, the named players. The dual-use capability. The four control surfaces. The concentration tendency baked into compliance-by-scale. The technically-true denial. The 2022 freeze as a real precedent.
"They harmonised it in 2025 in order to deceive the public while secretly building a CBDC."
That's intent you can't see and don't need. Convergent incentives, lobbying, and genuine policy fights produce the identical outcome — without a single villain.
The structure is the point, not a cabal. "BlackRock runs it" is the gravity well to avoid: it's a class of operators and a design, and the design is what does the work.
It is not a flat inevitability. It's contested, and incomplete.
Honesty about what cuts the other way is what keeps this from reading as doom:
Cash-protection and "right to cash" laws are being pushed, including in the EU. Decentralised and privacy-preserving alternatives keep appearing. Rival blocs are building rails outside the dollar chokepoint — China's e-CNY, cross-border CBDC experiments, BRICS settlement — because everyone watched Russia 2022 and learned that access to Western rails is revocable. The G7's harmonisation and the anti-G7 rail-building are the same phenomenon seen from two ends: everyone now treats payment infrastructure as a weapon to either hold or escape.
And the efficiency is genuinely real. Faster settlement, fewer intermediaries, cheaper cross-border payments — that's why the thing has momentum, and why "it's all a scam" misreads it. The sharp version isn't "tokenisation is bad." It's: the same thing that helps you is the thing that can steer you, and the authority to use it is being legislated in a different building so the two debates never meet.
Sources
↓ Download the one-page brief (PDF)