Bank of England’s Bailey warns AI benefits may take time to emerge

Bank of England Governor Andrew Bailey wants everyone to pump the brakes on AI optimism. Not because he thinks the technology is overhyped, but because history suggests the really transformative stuff takes longer than anyone expects to actually show up in the economic data.

Bailey’s argument boils down to a pattern that economists have seen before: powerful new technologies often make things worse before they make things better. The initial phase demands heavy capital investment, workforce retraining, and organizational restructuring. All of that costs money and time. And during that transition, productivity growth can actually slow down.

The electrification playbook

Bailey categorizes AI as a “general purpose technology,” placing it in the same lineage as electrification and the internet. That comparison is instructive, and not in the way Silicon Valley pitch decks would prefer.

When factories first adopted electric power in the late 19th and early 20th centuries, productivity didn’t immediately surge. Companies had to redesign entire factory floors, retrain workers, and develop new management practices. The economic payoff was enormous, but it arrived decades after the technology itself was available.

The valuation question

Bailey’s comments carry a second, more pointed message directed at financial markets. He flagged concerns about equity valuations tied to AI, noting that while current stock prices are backed by actual earnings, those earnings depend on the assumption that AI-driven profit growth will be sustained over time.

That’s a meaningful distinction. A bubble typically describes prices detached from fundamentals. Bailey isn’t quite calling the current AI trade a bubble. He’s saying the fundamentals themselves rest on a forward-looking bet that could easily disappoint.

Bailey’s warning also extends to systemic risks that AI introduces into the financial system itself. Model concentration, where large portions of the industry rely on the same underlying AI systems, creates a new category of correlated risk. If a widely used model produces flawed outputs or suffers a failure, the impact wouldn’t be contained to a single firm.

Cyber threats represent another dimension of concern. As AI becomes more deeply embedded in financial infrastructure, the attack surface for malicious actors expands.

What this means for investors

Bailey’s perspective matters because the Bank of England sits at the intersection of monetary policy, financial regulation, and systemic risk monitoring. When the governor of a G7 central bank suggests that AI’s economic benefits might arrive on a longer timeline than markets are pricing in, portfolio construction should reflect that uncertainty.

For crypto markets specifically, Bailey did not reference digital assets in his AI remarks. But the broader message is relevant. Much of the recent enthusiasm around AI-crypto convergence, from decentralized compute networks to AI-powered trading protocols, rests on the same assumption Bailey is challenging: that AI adoption translates quickly into measurable economic value.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.