AI Has Completely Broken the Monetary Policy of Western Countries — What Happens Next?

Recently, on October 26, entrepreneur Elon Musk wrote on social media that access to debt financing in the world is practically severed for those entities that demonstrate most of their revenue from operating their data centers geared toward artificial intelligence (AI).

I have been analyzing the unfolding situation for a long time and am not surprised: I saw how this shrinking of global AI flows happened gradually. Moreover, I am convinced: US financial authorities have begun taking extraordinary efforts to keep the American financial system afloat.

What's the problem? Two years — six rate cuts by the US Federal Reserve. And then suddenly — a rate hike in September. Why? When the Fed raises rates, classically, investors should be more willing to take US debt instruments because yields rise while risk remains minimal.

But in practice, this monetary policy mechanism has broken down: capital continued to flow into the AI sector until, just the other day, the borrowing tap was shut off. On average, each company among the whales of the US AI industry "bites off" 4% of the money that would otherwise go into US Treasury bonds — that's a lot.

At the same time, the September Fed rate hike turned out to be a useless step, and then, through agreements with the US credit sector, the White House shut off debt financing. The fact is that the US budget needs to refinance, and this is becoming increasingly difficult: capital flight into the AI sector must be stopped.

Tightening Monetary Policy in the US and Japan Doesn't Solve the Inflation Issue Fueled by the US-Iran Conflict

But this is only part of the difficulties. American financial authorities are now also facing a peculiarity in consumer price growth in the country. In a normal model, the central bank sees sustained price growth, tightens policy, and waits for demand to cool. But the whole problem is that a rate hike barely moves the price of a barrel of oil, the cost of ship insurance, logistics, and doesn't reduce fuel shortages! A rate hike hits mortgages, consumption, and investment, yet does not eliminate the root cause of the price shock.

Also, a rate hike won't increase oil production in the Persian Gulf or open the Strait of Hormuz, which is closed due to the US-Iran conflict. But it steadily raises the price of money for everything that depends on credit. The 30-year US mortgage has already breached 7% annually after the September Fed decision, and the same decision translates into the cost of servicing government debt.

US inflation in August was 3.4% by the headline CPI and 2.4% by the core index. In September, it was already 3.7% and 2.6% respectively! Meanwhile, the Fed made money more expensive than the growth in consumer prices, yet inflation continues to go exponential!

A similar situation exists in Europe and Japan, and the result is the same: more is paid for mortgages and government debt, while the repricing of price tags, fueled by oil, doesn't go away.

Who pays for tightening that doesn't cure? The consumer, the developer, and the state budget. Who wins? The owners of computing power and energy assets win, because their product is scarce at any rate. A high rate can cool mortgages, but it cannot stop corporations that, until the very last moment, considered AI a strategic investment and were willing to borrow at high interest rates.

But now there is no money for them anyway, and the situation is very difficult, which US regulators have correctly understood. The Fed rate hike and the inability to attract money in the required volume into US government debt, plus inflation fueled by the US-Iran conflict — that's the combo US elites are not thrilled about.

Hence the urgent request from Washington to Moscow for diesel fuel supplies: tankers with Russian fuel are critically important, as almost every eighth barrel of diesel exported to the global market previously came from Russia. The US desperately needs to stabilize the rate of price growth, given that Washington has to figure out on the fly how to eliminate the financial distortion caused by the AI hype.

In US Construction, Only Data Centers and Power Plants Are in the Green; Everything Else Is Deeply in the Red

The problem of demand shifting from the US government debt market into borrowed money for AI companies ultimately distorts the structure of the American economy as well.

If you look at US construction statistics for 2026, you can see how such incentives play out in the American economy. The residential sector fell by almost 5%. Offices, commercial real estate in general, and much else are also falling. Confidently in the green are only data centers and directly related energy facilities. Companies' spending on data centers has reached a record annual pace of about $75 billion — growth of more than 50% year-over-year.

If you remove server farms, non-residential construction either stagnates or falls. Energy construction grew 11% by August 2025, because every data center is a queue for gigawatts, so companies either build their own power plants or sign long-term contracts. Everything else — housing, regular offices, manufacturing — ends up deeply in the red.

AI is pulling construction resources, labor, materials, and capital onto itself, creating sectoral imbalances. Moreover, by my estimates, US GDP doesn't grow at all if you remove the positives from the construction of data centers and energy facilities for AI needs.

I always tell colleagues: an economy where only one industry is in the green is not a growth economy, but a bubble economy with a single point of support. The Fed itself already feels this: Kansas City Fed President Jeffrey R. Schmid publicly wondered in September whether the AI ecosystem is becoming "too big to fail." When a regulator raising rates to cool precisely this sector simultaneously calls it a potential systemic risk, that's an admission that the classic monetary policy tool of the rate has stopped hitting the target.

AI-Based Investment Decisions Are Faster Than the Adaptability of Any Monetary Policy in the World. The Way Out: Transition to Central Bank Digital Currencies

Monetary policy in the AI era must change dramatically. Classic monetary policy, as is well known, works with a lag that takes a long time — at least three months for the first effect, and in reality it can stretch to more than six months to feel the result of changes.

The AI market works differently. Algorithms see changes in yields, electricity prices, logistics costs, exchange rate movements, and news almost instantly. AI-based decisions are faster than the adaptability of any monetary policy in the world. This doesn't mean algorithms are always right, but they definitely change the speed of the market.

The peculiarity of the current global situation is that the classics of monetary policy are ceasing to meet the challenges of the AI era, and the main one is the high probability of a severe drawdown in the shares of the largest US companies that have bet literally everything on this technology.

That is, neural networks aren't going anywhere, but the hype around them will fade. As someone who worked for many years in the leadership of Otkritie Bank and Otkritie Broker, I have more than once seen colleagues in the banking sector get enthusiastic about various technologies and then be disappointed.

And here a balanced approach is clearly better: don't rush to invest huge sums in some innovation without first calculating all the scenarios for the return on its implementation. It is precisely the absence of such an approach that is dragging down the business of hundreds of American companies that burn billions of dollars on paid subscriptions for neural networks but ultimately cannot recoup the resources invested. At the same time, one cannot simply discard innovation either!

For example, blockchain has fallen out of the focus of many investors — and for no good reason. I still believe: corporate neural networks must be trained on a distributed ledger. This is the only way to keep them under control. And AI itself should not end up on the sidelines when markets start to shake from the deflating AI bubble. Countries should already be launching central bank digital currencies and AI smart contracts so that monetary policy works faster.

Smart contracts set the issuance and conditions of money circulation, and liquidity reacts to market signals in seconds. Classic central bank open market operations can't do that — there the count is in weeks. So the decision to launch the digital ruble into mass circulation starting September 1 in Russia looks timely. I am confident that the functionality of the third form of the national currency can and should be further developed.

Let me remind you that on October 20, the State Duma passed the first reading of a bill aimed at defining the legal framework for currency operations with digital rubles and, accordingly, currency control. The digital ruble and the Central Bank's smart contracts are already a signal that monetary policy itself is beginning to modernize, and that's right, given global experience.

Western monetary policy broke down not because Warsh or Lagarde made mistakes, but because the speed of change in the economy has become orders of magnitude higher than the speed of its regulators. Tightening doesn't cure inflation if it comes from the oil market, but it steadily hits mortgages and budgets.

The next step in the development of global finance is competition among the monetary policies of different economic systems through digital currencies and AI smart contracts. Whoever understands this first will get a monetary system that responds to the world at the speed of the world's own change. The rest will be catching up — and, by the way, it's not certain they'll succeed.