Stablecoins, AI and innovation: where is the handbrake?

Yesterday, there were a few headlines that caught my attention as they are all related.

The first two were about stablecoins and whether you can trust them:

Bank of England and Financial Conduct Authority’s approach to joint regulation of systemic stablecoin issuers

BIS report flags stablecoins' flaws, dollarization risk

For the past few years, the crypto industry has been telling us that stablecoins are the future of money. Central banks have been responding with a rather different message: perhaps, but only if they behave like money.

This week illustrated that growing divide perfectly.

The first development came from the UK, where the Bank of England and Financial Conduct Authority (FCA) published their long-awaited blueprint for jointly regulating systemic stablecoins. At almost exactly the same time, the Bank for International Settlements (BIS) reminded everyone why regulators remain so cautious, warning that stablecoins still suffer from fundamental weaknesses and could accelerate the global dominance of the US dollar.

The UK’s approach is actually rather elegant with a proposal that recognises the FCA should not treat every stablecoin the same, and so regulators are creating a two-speed regime.

Most stablecoin issuers will sit under the FCA, much like other financial firms. However, once a stablecoin becomes large enough to matter for financial stability, and HM Treasury designates it as systemic, the Bank of England joins the party. The FCA continues looking after consumer protection, conduct and market integrity, whilst the Bank focuses on prudential resilience, operational robustness and systemic risk. In other words, one regulator looks after customers whilst the other looks after the financial system.

That is exactly how financial regulation should work.

The objective is not to kill innovation. It is to recognise that a stablecoin with a few thousand users is fundamentally different from one processing billions of pounds every day. Once a digital token starts looking like critical payments infrastructure, it deserves to be regulated like critical payments infrastructure.

The UK has also softened some earlier proposals after industry feedback. Capital requirements have been reduced, reserve rules made more commercially workable and the emphasis has shifted from imposing arbitrary limits towards ensuring resilience, redemption and trust.

Then came the BIS. If the Bank of England is cautiously optimistic, the BIS remains deeply sceptical. Its latest comments repeat an argument it has made consistently for several years. Stablecoins are useful technology, but they are still poor money. They depend upon private issuers rather than sovereign institutions. They remain vulnerable to runs if confidence disappears. Different regulatory regimes create opportunities for arbitrage. Most importantly, if dollar-backed stablecoins become the world’s preferred digital currency, countries with weaker currencies could experience accelerated dollarisation, reducing the effectiveness of domestic monetary policy and increasing capital flight during periods of stress.

That last point deserves far more attention than it receives.

Around 99% of the global stablecoin market is denominated in US dollars. Every new wallet holding digital dollars instead of local currency quietly strengthens the dollar’s global influence.

Washington may not have planned it this way, but stablecoins have become one of the most effective instruments of American monetary power ever created. They export the dollar without exporting a bank. That’s the whole reasoning behind the GENIUS Act imho.

Viewed through that lens, the UK’s regulatory framework becomes more understandable. The Bank of England is not simply writing crypto rules. It is protecting the integrity of sterling whilst allowing innovation to flourish.

The message is straightforward: if private digital money wants to become part of Britain’s financial infrastructure, it must meet the same standards of resilience, governance and operational robustness as every other critical payment system.

This is why the debate about stablecoins is no longer really about crypto. It is about who issues money, who controls payments and who ultimately governs the digital economy. Technology companies see programmable cash. Central banks see monetary sovereignty. Banks see competition for deposits. Consumers simply want something that works.

The irony is that everyone may be right.

Stablecoins almost certainly will become a permanent part of the financial landscape. The question is whether they evolve into trusted public infrastructure or remain private digital chips circulating around the edges of finance. The UK is betting they can become the former through regulation whilst the BIS is warning that, without international coordination, they may become the latter.

Then add artificial intelligence to the mix and the regulatory anxiety becomes even clearer as evidenced by a speech by Sarah Breeden, Deputy Governor of the Bank of England:

Sarah has warned that AI is reshaping finance at speed, particularly through autonomous agents operating in cyber security, markets and payments. Her central point is simple: the next technology surprise must not become the next financial stability test.

This connects directly to the stablecoin debate.

Stablecoins make money programmable. AI agents make decisions programmable. Put the two together and you create a financial system where machines can hold value, move value, trade value and potentially panic at machine speed.

That is why the Bank of England is now talking about circuit breakers and even AI “kill switches” for autonomous trading systems. The concern is not that one rogue bot makes a bad trade. The concern is that thousands of similar models respond to the same signal in the same way at the same time, amplifying volatility rather than absorbing it.

In old finance, markets crashed because humans panicked and now, in new finance, markets may crash because machines agree.

That is a very different problem. Human panic is messy, emotional and uneven. Machine panic can be synchronised, instantaneous and global. Add stablecoins, tokenised assets and real-time payment rails, and you no longer have a financial system that waits for settlement. You have one that moves before regulators have finished reading the dashboard.

This is the real issue.

The debate is no longer just about stablecoins, AI or crypto. It is about the emergence of autonomous finance: private digital money, intelligent agents, programmable markets and payment systems that operate continuously.

The Bank of England and FCA are trying to regulate systemic stablecoins because private money can become infrastructure and Sarah is warning about AI agents because private decision-making systems can become market infrastructure. Meanwhile, the BIS is warning about dollarisation because private digital currencies can become geopolitical infrastructure.

Different topics. Same story.

Finance is becoming machine-readable, machine-executable and machine-governed. That’s a fact and so, the regulators are not trying to stop that future. They are trying to work out where the brakes, seatbelts and emergency exits should be before the vehicle reaches full speed.