Too Digital to Fail (Part Four)

Digital money doesn't have a passport.

For millenia, money has been organised around geography. The dollar belongs to America, sterling to Britain, the euro to Europe and the yen to Japan. Banks are licensed nationally, central banks provide national monetary anchors, governments guarantee deposits nationally and regulators supervise institutions largely through national legal frameworks.

Digital currencies break that architecture.

Bitcoin does not know whether its owner is sitting in London, Lagos or Los Angeles. A dollar stablecoin does not stop at the American border. A wallet does not need to understand whether the person sending money is in Manchester and the recipient is in Mumbai. The blockchain simply records that value moved from one address to another.

That sounds obvious, but its implications are profound because we are creating global money while retaining national regulation.

This is exactly the tension emerging with AI and cloud. Cloud infrastructure is global. AI models are global. Cyberattacks are global. Crypto networks are global. Stablecoins are global. Meanwhile, governments, central banks, deposit guarantee schemes, insolvency regimes and financial regulators remain overwhelmingly national.

That creates an increasingly uncomfortable question:

Who is responsible when global money creates a national crisis?

Stablecoins are where this gets particularly interesting

Stablecoins take the currency of one country and make it digitally global.

A dollar stablecoin turns the US dollar into something that anyone with internet access and an appropriate wallet can hold and transfer around the world. The Federal Reserve does not need to open an account for that person. A local bank does not need to provide the dollar deposit. The user gets economic exposure to dollars through a private digital instrument.

That makes stablecoins far more important than another payments innovation. They are becoming a new distribution system for currencies.

The strongest currencies therefore gain an enormous advantage. Someone living in a country with inflation, currency controls or an unstable banking system has an obvious incentive to hold digital dollars rather than the domestic currency.

That creates an opportunity for America because dollar stablecoins extend the reach of the dollar deep into the digital economy. It creates a much more uncomfortable problem for everyone else because digital dollarisation no longer requires suitcases of cash, offshore bank accounts or sophisticated financial arrangements.

It requires a smartphone and, in doing so, it turns currency competition global. This is where CBDCs become part of the same story.

Central banks originally discussed CBDCs largely as domestic payment instruments, but the strategic question is increasingly international. If dollars, euros, yuan and privately issued stablecoins become digital, programmable and globally accessible, currencies increasingly compete on technology as well as economics.

Which currency has the best infrastructure? Which settles fastest? Which is accepted by the most platforms? Which works across the largest number of wallets? Which has the deepest liquidity? Which currency will AI agents choose when they negotiate with each other?

That last question becomes particularly important as humans have national identities and habits. Machines do not.

An autonomous purchasing agent in Britain buying components from an autonomous supplier in Korea has no emotional attachment to sterling or won. It wants the currency or token offering the lowest transaction cost, deepest liquidity, fastest settlement and lowest counterparty risk.

AI agents will choose money algorithmically.

That introduces a completely different form of currency competition because machines will constantly optimise between currencies, stablecoins, tokenised deposits and other digital assets ... and then the borders really start disappearing

Imagine millions of AI agents continuously managing corporate and personal money.

They monitor interest rates, exchange rates, counterparty risk, political developments, bank creditworthiness and market liquidity. They automatically move funds towards whatever instrument offers the best combination of safety, liquidity and return.

A political crisis develops in one country. The machines detect rising risk. They do not wait for the central bank governor's press conference. They do not watch television. They do not panic. They optimise.

Domestic currency becomes dollars. Bank deposits become stablecoins. Stablecoins move between issuers. Deposits move into CBDCs. Assets cross blockchains and jurisdictions automatically.

The capital flight that once took days or weeks takes seconds.

Your existing AI argument is that thousands of institutions using similar models can simultaneously sell the same assets, creating a feedback loop of falling prices, rising measured risk and further selling.  Digital currencies add another dimension:

AI doesn't just automate the investment decision. Digital money automates the escape route and that changes the meaning of a bank run. The traditional bank run was physical. The modern bank run is digital. The next bank run is global and autonomous.

Imagine an AI treasury system determines that Bank A has become risky. It moves corporate deposits into a stablecoin. Other agents detect the same signals and do the same thing. Those stablecoins are then moved into tokenised US Treasury securities or another currency.

The money has not simply left the bank.

It has left the banking system, crossed the border and changed currency almost simultaneously.

National regulators then face a problem they were never designed to solve. The central bank can provide liquidity to its domestic banks, but it cannot order a decentralised blockchain to stop. The government can guarantee domestic deposits, but it cannot stop autonomous agents choosing digital dollars instead. Capital controls become considerably harder when capital exists as globally transferable tokens.

The central problem therefore stops being simply too big to fail or even too interconnected to fail. It becomes: too global to control.

There is an enormous opportunity on the other side, however. as this should not become another doom article because the same borderlessness solves some of finance's oldest problems.

Cross-border payments remain unnecessarily slow, expensive and complicated precisely because the financial system is organised around national infrastructures connected through layers of correspondent banking, messaging, compliance, reconciliation and settlement.

Digital currencies attack that architecture directly.

A genuinely interoperable world of regulated stablecoins, tokenised deposits and CBDCs offers instant cross-border settlement, dramatically lower remittance costs, 24/7 global commerce, programmable trade finance and autonomous machine-to-machine payments.

For banks, the opportunity therefore shifts from moving money between countries towards providing trust around money moving globally.

The challenge is that the financial system is becoming global faster than its governance.

Cloud has already globalised the infrastructure underneath banking. AI is globalising the intelligence operating across that infrastructure. Digital currencies globalise the money itself ... and yet regulation remains national. Central banks remain national. Deposit insurance, fiscal policy, bank resolution and political accountability remains national.

That is the structural contradiction.

The opportunity is a genuinely global financial system in which money moves as easily as information. The risk is a genuinely global financial system that no national authority fully controls.

• Cloud removes the technological borders.
• AI removes the decision-making borders.
• Digital currencies remove the monetary borders.

As these three things come together, we have created something finance has never had before: a global, autonomous, machine-speed financial system governed by institutions designed for a national, human-speed world.

That, to me, is the real risk and extraordinary opportunity.