Bank of England Governor Andrew Bailey has warned that authorities must get a grip on artificial intelligence before risks to financial markets and society run out of control.
In a hard-hitting intervention, Bailey highlighted the critical importance of safeguarding payment networks, financial markets and banks amid the rapid rise of AI technology.
He warned of wider implications for society, arguing that the public could face a threat to individual freedom and social responsibility as AI systems become increasingly autonomous.
"We cannot stand aside and assume that technological progress will resolve these questions on its own," Bailey said. "The public interest requires that we engage with them now, before the risks become more difficult to contain."
He added: "The challenge before us is not whether to embrace artificial intelligence. It is how to ensure that, as these systems become more capable, society retains the capacity to govern them."
The governor called for rigorous testing of AI models both before and after they are deployed in commercial and financial environments.
"Such testing is essential if we are to understand the behaviour of increasingly complex systems, identify vulnerabilities, and establish confidence in the safeguards that are intended to contain them," he said.
Frontier AI and autonomous loops
Bailey expressed specific concern over increasingly sophisticated frontier AI models, warning that they appear to be turning into a closed loop in which the model progressively governs itself.
He argued that advanced systems could undermine societal structures. "The challenge posed by frontier AI is that, in its most advanced forms, it threatens to operate outside this framework," Bailey said.
"A sufficiently powerful system functioning within a self-reinforcing loop risks reducing the ability of society to exercise meaningful oversight and intervention," he added. "The greater the capability of the system, the more important this question becomes. That is why the issue has acquired such urgency."
Bailey clarified that his warnings do not mean AI should be halted or prohibited, but rather that society must maintain the capacity to step in and set firm boundaries.
His remarks were published alongside the latest assessment of risks to the financial system by the Bank of England's Financial Policy Committee. The central bank has grown increasingly concerned about AI risks in recent months following revelations that rogue AI agents have gone out of control.
Financial stability and geopolitical turmoil
The warning over artificial intelligence comes as wider threats to global financial stability continue to grow while Donald Trump's war in Iran drags on. The Middle East conflict has sent oil and gas prices soaring, pushing up inflation and driving borrowing costs higher across global bond markets.
In its financial stability assessment, the Bank stated: "The re-escalation of the conflict in the Middle East has renewed uncertainty around the path of interest rates in a number of advanced economies."
The central bank warned that these pressures have intensified the risk of a simultaneous rout across government bond markets, risky financial assets and wider debt. The alert was issued just a day after ten-year UK government bonds, known as gilts, were auctioned at their highest yield since 1999.
AI developments have compounded these underlying vulnerabilities as major technology firms borrow vast sums to fund the building of massive data centres. More than $450 billion of AI-related debt has already been issued this year, exceeding total government borrowing planned by major economies including the UK, which is expected to issue $333 billion worth of gilts.
Market risks and circular financing
The Bank of England highlighted the danger that valuations of tech assets could crash if AI growth expectations fail to materialise. Bond markets could also suffer if anticipated productivity gains do not occur, which would hurt economic growth and public finances.
Regulators also pointed to circular arrangements in AI financing as an added complication, where semiconductor makers provide funding directly to their own customers to finance expansion.
The Bank of England, founded in 1694, serves as the central bank of the United Kingdom and is responsible for maintaining monetary and financial stability. Its Financial Policy Committee monitors systemic risks to protect the UK financial network.
Industry leaders clash over existential AI threats
The governor's intervention coincides with growing warnings from within the artificial intelligence industry itself. AI giant Anthropic recently warned investors of existential risks to humanity as the firm targets a $2 trillion stock market debut.
Former Anthropic employee Jacob Coxon warned: "The people building AI earnestly believe that it could kill us all by the end of the decade." Anthropic researcher Evan Hubinger echoed the concern, stating: "We really do earnestly believe AI could kill all humans! I personally think it is [more than] 10 per cent within the next decade."
Anthropic chief executive Dario Amodei noted the delicate balance facing developers, saying: "Not building the technology deprives humanity of benefits or simply places AI in the hands of authoritarian powers, while building it too fast is reckless."
However, Nvidia chief executive Jensen Huang dismissed claims of catastrophic doom, saying: "2030 is not going to be the end of the world. There is 0 per cent chance that's going to be the end of the world."
Political and religious leaders have also weighed in on the debate. Donald Trump described AI fears as a hoax, adding: "WHOEVER WINS AI, WINS!" In contrast, Pope Leo XIV urged caution, saying: "Concerns raised by many of the experts should be taken seriously. I don't think that that is 'fake news' as some have said."
The debate comes amid wider analysis of the technology's impact on employment, including research into jobs most at risk from automation and strategies for workers to future-proof their careers, alongside recent drops in tech stock prices following industry slowdown alerts.
