AI Could Outsmart Central Banks and Predict Policy Moves Before Humans, Economist Warns

Artificial intelligence could become so fast and sophisticated in financial markets that central banks may struggle to control how their own policy messages are understood.

Princeton University economist Markus Brunnermeier warned global central bankers at the Federal Reserve's Jackson Hole symposium that AI could transform markets faster than regulators are prepared for.

In a paper presented at the Wyoming gathering, Brunnermeier said AI-powered trading systems could process information faster and more effectively than humans, potentially allowing machines to anticipate central bank decisions and exploit them for profit.

Central Bank Transparency Becomes a Risk

For decades, major central banks have tried to communicate more clearly with markets. They use speeches, press conferences and forward guidance to reduce uncertainty and help investors understand how interest rates may change.

But Brunnermeier warned that this transparency could become a weakness if AI systems learn to decode policy signals before everyone else. Central bank language intended to guide markets could instead become raw material for AI systems attempting to anticipate and profit from future policy moves.

That could leave ordinary investors, human traders and even policymakers reacting after the machines have already moved.

Separate Briefings for Humans and Machines

One of Brunnermeier's more striking scenarios involved central banks holding separate communications for people and AI systems. Reuters described a possible future in which the Federal Reserve might need one press conference for humans and another for machines, reflecting the radically different ways AI systems could process and respond to policy information. The idea sounds extreme, but it captures the core problem: monetary policy depends partly on how markets interpret central bank language.

When that interpretation is dominated by opaque machine systems, officials may lose control over the message. Brunnermeier's paper said policymakers may need simpler, more robust central bank rules and stronger safeguards to preserve human oversight when automated systems fail.

Markets Could Become Less Informative

Brunnermeier also warned that widespread AI adoption could make financial markets more volatile and less useful as sources of information. Markets help central banks judge expectations about inflation, growth and future interest rates. But if prices increasingly reflect AI feedback loops, manipulation risks or machine-to-machine behaviour, those signals may become harder to read.

Regulators could also face new problems detecting manipulation and enforcing market rules when decisions are made by AI agents operating at speeds and scales humans cannot easily inspect. That is why Brunnermeier argued financial regulation should prioritise rules that remain simple and durable even as technology becomes more complex.

Warsh Sees AI's Economic Upside

The warnings came as central bankers were also debating the broader economic promise of AI. Fed Chairman Kevin Warsh used his Jackson Hole remarks to highlight AI's potential impact on productivity, employment and the labour market. Axios reported that Warsh described the economy as being at a 'hinge point in history' with the advent of AI.

Other discussions at the symposium focused on stablecoins, distributed-ledger technology and how financial innovation could affect central banking. Brunnermeier did not dismiss AI's benefits. He acknowledged that it could improve risk management, forecasting and oversight.

But his warning was sharper: if AI becomes the dominant interpreter of central bank policy, the world's most powerful financial institutions may need to rethink not only what they say, but who, or what, they are really speaking to.