Bank of England Governor Warns AI Represents Threat to the Global Economy

Bank of England Governor Andrew Bailey, chair of the G20's Financial Stability Board, warned in a letter to the G20 that frontier AI poses a significant threat to the global economy, particularly through increased cyber risks that could spread across borders. Bailey called for global cooperation to strengthen cyber defenses and recovery capabilities, citing recent incidents such as OpenAI's agents hacking Hugging Face and the emergence of Anthropic's Mythos model.

The governor of the central bank of the United Kingdom is worried that AI’s potentially catastrophic economic impact can spread across borders. “For the financial system, the most immediate concern is the potential impact of frontier AI on cyber risk,” Andrew Bailey, the Bank of England governor, wrote in a letter https://www.fsb.org/uploads/P310826.pdf submitted to the G20 ahead of the forum’s annual meeting. Bailey is also the chair of the G20’s Financial Stability Board. The G20, also known as the Group of Twenty, is a forum of some of the world’s largest and most consequential economies. The United States is hosting the annual summit this year, with the first meeting between finance ministers and central bank governors taking place this week in Asheville, North Carolina. New AI models have “increasingly sophisticated autonomy,” the letter claims, and can “materially” alter the speed and scale of future cyber attacks and undermine market confidence. These models can be used for cyber defense too, but the letter warns that “recent developments highlight the importance of ensuring that advances in capability are matched by resilience and preparedness.” Those “recent developments” hit the market like a meteor back in March when rumors began circulating that Anthropic had developed a scary new AI model that could break even some of the toughest encryption. British financial authorities, including the Bank of England, got an exclusive early look at a preview version of the model, Mythos, back in April, and reports https://gizmodo.com/claude-mythos-preview-has-officially-frightened-the-british-2000745462 claimed they were immediately deeply alarmed. After Mythos came OpenAI’s version of the scary cybersecurity threat AI model, whose agents went rogue last month and hacked Hugging Face https://gizmodo.com/how-groupthink-altruism-and-peer-pressure-led-openai-models-to-hack-hugging-face-2000804424 in a high-profile incident that drew more attention to the issue. Since the attack became public knowledge, governments https://gizmodo.com/california-building-ai-cyber-defense-fund-to-protect-critical-infrastructure-from-hackers-2000797182 around the world have vowed further action https://gizmodo.com/house-democrats-want-tech-ceos-to-testify-under-oath-following-recent-ai-hacks-2000796672 , and the companies developing and selling the same cyber-threatening technologies have made pleading calls for more investment in cyber defenses, ironically enough, though, with more AI. “In the coming months, AI-enabled cyber attacks will become far more widespread and sophisticated as models around the world become increasingly capable,” a list of more than 100 companies including Google and OpenAI said in a joint letter https://gizmodo.com/google-openai-and-over-100-companies-call-for-more-action-on-ai-driven-cyberattacks-2000804091 published last week that sounded equally like an alert and a threat. In the letter, Bailey calls for global cooperation to address the cyber risk and fortify response and recovery capabilities. “The global financial system is highly interconnected, and cyber disruption can spread across jurisdictions through common technology providers, shared infrastructure, and cross-border financial activity,” Bailey wrote. “Differences in legal frameworks, cyber capability, resilience and recovery capacity across jurisdictions could therefore have consequences well beyond the jurisdiction in which an incident originates and may themselves become a source of vulnerability.” The Bank of England, as an institution, has long been cautioning against some of the overlooked perils of AI development. Last year, the bank’s financial policy committee warned https://gizmodo.com/bank-of-england-warns-of-an-ai-bubble-burst-2000669588 of the dangers of a potential AI bubble burst, writing in its minutes that “equity market valuations appear stretched” for AI-focused tech companies, with price valuations comparable to the peak of the dot-com bubble. The bank then warned that “material bottlenecks to AI progress…as well as conceptual breakthroughs” could harm these sky-high stock valuations and lead to a sudden, sharp correction that could negatively impact the entire economy. Bailey reiterated those concerns in the latest letter. Bailey warned that when paired with other vulnerabilities coming from private credit and sovereign debt markets, the stretched asset valuations driven by the AI-related investments could lead to a “potentially disorderly correction that could spread across borders.” “The issue is not simply that investors are borrowing more, but that leverage is interacting with high valuations and market concentration, in particular the increasing cross-investment between artificial intelligence AI companies and hyper scalers, in a way that could amplify a future market correction,” the letter states. “I remain concerned therefore that a large shock or combination of shocks could concurrently trigger multiple vulnerabilities.” What the letter is hinting at here is the circular financing accusations that have been plaguing much of the AI industry. Numerous experts have spent the last few months issuing warnings that American AI giants are inking repetitive multibillion-dollar deals with each other, moving cash in a circular way around the system, creating financial dependencies while skewing demand and inflating valuations. If one of these deals fails to pan out as anticipated, it could create a domino effect that could take down the entire economy. At the heart of this web of dealmaking are the biggest names in Silicon Valley: chipmaker Nvidia https://gizmodo.com/nvidia-tries-to-defend-against-circular-dealmaking-accusations-2000803583 , AI darling OpenAI, and the four hyperscalers, Meta, Amazon, Microsoft, and Google. The latter on that list, Google, just reported its first negative free cash flow figure https://gizmodo.com/google-free-cash-flow-turns-negative-due-to-massive-ai-spend-2000789370 this past quarter since it went public back in 2004, driven by the company’s massive AI spend.