The people in charge of keeping the global economy from flying into a mountain just got told that their instruments might not work anymore. According to a new paper from the Bank for International Settlements, the AI boom is actively scrambling the economic indicators that central bankers have depended on for decades. Sleep tight.
What the BIS Actually Said
The Bank for International Settlements, which is essentially the central bank for central banks and operates out of Basel, Switzerland, published a paper finding that artificial intelligence is blurring the signals that monetary policymakers use to do their jobs. We're talking about inflation readings, employment figures, financial stability metrics — the whole dashboard.
The core problem, as Axios reports, is that AI is hitting both sides of the economy at once. Supply and demand. Structural change and cyclical change. All simultaneously. That's a problem because most central banking frameworks were built on the assumption that you could at least identify which kind of problem you were dealing with before you started pulling levers.
Why 'Rules of Thumb' Are Kind of the Whole Job
Here's something the financial press doesn't explain enough: central banking runs on heuristics. The Taylor Rule. The natural rate of unemployment. The yield curve as a recession predictor. These are not laws of physics. They are educated guesses that worked consistently enough over enough decades that policymakers started treating them like laws of physics.
When the BIS says AI is undermining the rules of thumb central bankers rely on, it is saying something genuinely alarming. It means the Fed, the European Central Bank, the Bank of England and everyone else are trying to read a map that may or may not reflect the territory anymore. Rate decisions that affect mortgages, business loans, and whether people keep their jobs are being made with instruments that one of the most credible bodies in global finance is now flagging as unreliable.
Both Sides of the Economy, at the Same Time
The specific complexity the BIS paper identifies is that AI is not a simple shock. A hurricane hits demand. A new oil field hits supply. Policymakers know how to think about those. AI, per the BIS findings reported by Axios, is doing both things simultaneously while also scrambling the distinction between short-term cyclical fluctuations and longer-term structural shifts in the economy.
That distinction matters enormously. If inflation is rising because of a temporary demand surge, you raise rates to cool things down and eventually you stop. If inflation is rising because the entire productive structure of the economy is reorganizing itself around a new technology, raising rates might do very little except hurt people who borrowed money. The BIS is essentially saying we may not be able to tell which situation we're in. That's not a small thing to admit.
Nobody Actually Knows How Big This Is Yet
The AI productivity boom is real, but it is arriving unevenly and faster than anyone's measurement tools were designed to track. The Bureau of Labor Statistics was not built for a world in which millions of workers are quietly offloading portions of their jobs to software that costs forty dollars a month. GDP accounting was not designed to capture value that gets created and consumed without a transaction anyone records.
Meanwhile the labor market data that Jerome Powell squints at every month is itself a product of survey methodology designed in an era when the nature of work was more legible. If AI is reshaping what jobs actually involve faster than surveys can capture, the unemployment rate might look fine while the actual story of the labor market is something considerably weirder. The BIS paper is a high-level warning, but it is pointing at a very concrete problem.
The Timing Is Not Great
Central banks are only now getting their footing after the worst inflation episode in forty years, a period that already exposed significant blind spots in how economists model price dynamics. The Fed famously called inflation transitory for longer than it should have. The BIS paper lands at exactly the moment policymakers would most like to believe they have their analytical tools back in working order.
They may not. Axios reports the BIS paper is framing this as an ongoing, evolving challenge, not a problem with a clean solution on the horizon. The AI investment wave is still accelerating. The economic effects are still compounding. The indicators central banks watch are going to keep getting blurrier before anyone figures out how to sharpen them.
The Dingo Take
You are supposed to feel reassured that the adults in charge of global monetary policy have sophisticated models and rigorous data and years of institutional expertise. The Bank for International Settlements just published a paper strongly implying those models are operating in conditions they were not designed for, reading data that may not mean what it used to mean, during a technological transition that is affecting every variable at once. The adults are not panicking, which is professionally appropriate. But they are issuing formal warnings through the central bank for central banks, which is the institutional equivalent of the pilot telling the flight crew in a very calm voice that the altimeter is behaving strangely.
The maddening part is that this is no one's fault in a simple sense. AI is not a policy choice gone wrong. It is a technological wave hitting an economic system whose measurement infrastructure was built for a different world. But 'no one's fault' does not mean 'no consequences.' The consequences land on ordinary people: on the family whose mortgage rate got set too high because the Fed misread an inflation signal, on the workers who got written off as 'employed' in a survey that couldn't see how their work was quietly being automated underneath them.
The BIS is not crying wolf. This is the most credible financial institution in the world quietly telling anyone paying attention that the framework is straining. The question is whether the people making decisions that affect billions of lives will take that seriously, or whether they'll keep citing the old rules of thumb because the old rules of thumb are what they know how to use. History suggests we'll find out the hard way.
