The UK’s top financial watchdog just admitted it’s losing ground to the machines. FCA Chief Executive Nikhil Rathi used a speech on June 24 to warn that regulators are locked in an “arms race” with artificial intelligence in financial services, one where the technology side has a serious head start.

The core problem is straightforward: AI agents can adapt and evolve within weeks or months. Regulatory frameworks, built for a world of quarterly reports and annual reviews, operate on a fundamentally different clock.

The scale of the problem

Rathi cited studies showing over 80% of financial services firms are already incorporating AI into their operations. Rathi delivered his warning at a techUK event focused on generative AI. The speech landed months after the House of Commons Treasury Committee published a report in January 2026 that essentially told regulators their “wait-and-see” approach was a dangerous gamble.

Bank of England Deputy Governor Sarah Breeden has separately cautioned about autonomous AI agents operating in trading and payment systems. Her specific concern: these agents could trigger a “market meltdown” if left unchecked.

The UK currently relies on existing regulations like the Senior Managers and Certification Regime to govern AI use in finance.

A framework that wasn’t built for this

The FCA has maintained what it calls a “technology-neutral” regulatory stance. The Treasury Committee’s January 2026 report poked significant holes in that logic. When the tool itself can autonomously make decisions, adapt its strategies, and interact with markets at speeds no human can monitor, the distinction between tool and actor starts to blur.

The Mills Review on retail financial services, also initiated in January 2026, examined how technology is reshaping consumer finance. Rathi’s speech was a direct acknowledgment that the current toolkit is insufficient. The FCA needs enhanced regulatory tools and powers, and it needs them before the AI systems it’s trying to oversee become too sophisticated to govern effectively.

What this means for investors

For firms heavily invested in AI-driven trading and financial services infrastructure, the calculus is complicated. The current light-touch approach means fewer compliance costs and more room to innovate. When regulation eventually does arrive, it could come in heavy and fast.

The Bank of England’s “market meltdown” warning from Breeden deserves particular attention from traders. If autonomous AI agents are concentrated in similar strategies, using similar data, and reacting to similar signals, the potential for correlated behavior during stress events is real.

With over 80% of firms already using AI, the technology itself is no longer a differentiator. Companies that proactively build compliance-ready AI systems may find themselves with a structural advantage when rules finally materialize.