When AI escapes control: the Bank of England’s warning goes beyond the markets.

  Articoli (Articles)
  Sarah Azzurra Spada
  09 October 2026
  3 minutes, 2 seconds

Artificial intelligence is entering markets with a now-familiar promise: a productivity revolution that can generate returns on any investment. The industry projects a simple, immediate image of a technology destined to transform the economy, fuelled by unprecedented capital. The five largest U.S. tech companies currently account for roughly 30% of the S&P 500, representing the highest concentration in half a century, while global AI debt issuance reached $450 billion by early September, more than doubling the 2025 total. However, these figures do not just reflect a financial success; they also represent a structural fragility that the Governor of the Bank of England, Andrew Bailey, has chosen to explicitly address.

The first signal came in late August, in a letter sent to Finance Ministers and G20 governors gathered in Asheville, North Carolina. As Chair of the Financial Stability Board, Bailey warned that a “future market correction” could spread worldwide. What is concerning is not a single factor, but rather their combination: high valuations, market concentration, and cross-investments between AI companies and major infrastructure providers, all amplified by financial leverage. The Governor’s tone remains that of a central banker. He said, “You could see some correction of asset prices at some point” when the BBC asked him whether the bubble could burst. It is a prudent formulation that nonetheless leaves all possibilities open.

The most interesting part, however, is a different one. In the essay Frontier AI and the question of governance, published on the Bank’s website on September 30, Bailey shifted the discussion from costs to technology. The most advanced models, he noted, increasingly learn from their own outputs, creating a self-referential circuit that risks undermining society’s ability to exercise oversight. This is a crucial point, because the core question is not about regulation itself, but whether it remains possible to intervene and set boundaries that can be revised as technology evolves: the governor’s answer is an unequivocal “yes”. His next point is just as clear: regulation is not the starting point. What comes first is understanding, rigorous testing both before and after a model’s release, and credible points of intervention.

Why is a central bank addressing this issue? A clear example is cyber risk. Bailey wrote that frontier AI increases the scale and sophistication of threats to payments, market infrastructures, and banks. This is the most immediate concern for financial stability, and it is the same concern highlighted in the letter to the G20.

To fully understand Bailey’s position, it is worth highlighting a fundamental point. The governor is not calling for a halt to AI, whose potential benefits for research and productivity he acknowledges. His primary concern is a loss of control, even before a loss of value. Therefore, the AI bubble should not be viewed merely as a market risk, but also as a symptom of a technology advancing faster than the capacity to govern it.


Mondo Internazionale APS - Riproduzione Riservata ® 2026

Translated by Paola Ghisu

Share the post

L'Autore

Sarah Azzurra Spada

Tag

artificial intelligence Andrew Bailey CyberSecurity bank of england financial stability financial markets regulation AI