The Bank of England's governor has issued one of the most sobering assessments of artificial intelligence to emerge from a central banking institution this decade: that new AI models represent a genuine and escalating risk to the global financial system. The warning, which surfaced at the opening of September 2026, carries exceptional weight not only because of its institutional source, but because independent analysts have noted that it mirrors concerns previously raised by the economist who correctly foresaw both the 2008 financial crash and the Covid-19 economic shock — a forecasting record that commands serious attention in global policy circles.
The convergence of these two voices — one from inside the most storied central banking establishment in Europe, and one from a figure whose predictive track record has earned rare credibility — gives this moment an urgency that goes well beyond the routine hand-wringing about technology disruption. When the same warning emerges simultaneously from the financial establishment and from its most credible outside critic, market participants and policymakers alike would be unwise to dismiss it as noise.
What exactly are these AI models doing that has alarmed the Bank of England's leadership? The concern centres on systemic risk: the possibility that AI systems, increasingly embedded in trading, credit decisions, liquidity management, and risk assessment across interconnected financial institutions, could amplify shocks rather than absorb them. The very speed and scale at which AI operates — executing decisions across thousands of nodes simultaneously — means that a flawed model or a coordinated failure could propagate through the global financial system far faster than any human oversight mechanism could intervene. The 2008 crisis demonstrated how opacity and interconnectedness could turn localised problems into cascading global catastrophe. AI introduces both of those qualities at a new order of magnitude.
Yet even as regulators grapple with AI's systemic implications at the macro level, a separate but equally consequential problem is emerging at the infrastructure layer. Analysts have identified payments infrastructure as the real bottleneck constraining AI software's potential — a finding that reframes much of the current conversation about artificial intelligence in finance. The dominant narrative has focused on model capability: how intelligent the AI is, how accurately it predicts, how well it reasons. But intelligence without the ability to execute transactions is commercially inert. If the pipes through which money moves cannot keep pace with the speed, volume, and complexity of decisions that AI systems generate, then the technology's promise cannot be fulfilled.
This is a structural problem decades in the making. Legacy payments infrastructure — built in many jurisdictions during the 1970s and 1980s and only partially modernised since — was not designed for the real-time, always-on, globally fragmented payment flows that AI-driven financial services require. Batch processing cycles, correspondent banking chains, and settlement lags that were once acceptable inefficiencies now threaten to become binding constraints. The irony is striking: the financial industry is investing aggressively in state-of-the-art intelligence at the front end while still relying on antiquated plumbing at the back end. AI can make a lending decision in milliseconds; disbursing that loan across borders may still take days.
Compounding both challenges is a third dimension that may prove the most consequential of all for the long-term architecture of global finance: central bank independence is, by credible assessment, facing its biggest stress test of this century. Political pressures on monetary authorities have intensified across multiple major economies, and the institutional norms that have underpinned central banking orthodoxy since the 1990s are being openly questioned. If the institutions best positioned to regulate AI-related systemic risk and to modernise payments oversight are themselves under political siege, the governance gap widens precisely when it needs to close.
The intersection of these three forces — AI-driven systemic fragility, infrastructure bottlenecks, and weakening institutional independence — is not a coincidence of timing. They are mutually reinforcing. A central banking system under political pressure is less able to impose the regulatory frameworks necessary to govern AI in finance. An AI-saturated financial system operating over decrepit payments rails is more vulnerable to the kind of cascading failure the Bank of England's governor is warning about. And a payments infrastructure that cannot scale will limit the economic upside of AI just as its risks are peaking.
What This Means for the Industry
For financial institutions, the message is clear: the AI agenda cannot be decoupled from the payments modernisation agenda. Boards that are approving AI investment without simultaneously scrutinising their payments infrastructure dependencies are building on an unstable foundation. For regulators and policymakers, the Bank of England's warning should be treated as a call to accelerate international coordination on AI risk frameworks — not as a reason for unilateral national action that could fragment oversight further. And for those watching the political pressures on central banks, the defence of institutional independence is not an abstract constitutional matter; it is a prerequisite for effective management of the most consequential technological transition finance has ever faced. The warnings are converging. The window for measured, coordinated response is narrowing.
Written by the editorial team — independent journalism powered by Codego Press.
Top comments (0)