ECB calls for review of financial infrastructure due to AI and stablecoins

José Luis Escrivá, a member of the governing council of the European Central Bank, warned this week that the growing influence of artificial intelligence requires a complete reassessment of the resilience of global financial infrastructure. Additionally, he raised a critical question about the need to strengthen the role of central banks as a last line of defense against systemic risks associated with the proliferation of cryptocurrency stablecoins, which are increasingly competing with the traditional banking system.

Why central banks worry about AI

Artificial intelligence is rapidly being integrated into financial systems worldwide. Major banks use machine learning for automated equity and currency trading, credit risk management, and real-time analysis of client behavior. By themselves, these applications increase efficiency, but they create new vulnerabilities.

The problem is that large-scale AI deployment could create systemic threats that traditional regulators are not equipped to track. This includes algorithmic trading failures (when one algorithm failure triggers a wave of sales) and the possibility of unpredictable interactions between AI systems from different banks operating in parallel. Escrivá emphasized that the current infrastructure was designed long before the era of large multi-parameter AI models like GPT and is not ready for the scale of potential impact.

If such an algorithm malfunctions or encounters an unforeseen market situation, the consequences could quickly extend beyond a single institution and affect the entire market. Previous crashes (2008, volatility spike in 2020) demonstrated how quickly systemic risks spread.

The stablecoin threat

A separate source of concern is the rapid proliferation of stablecoins (cryptocurrencies pegged to the US dollar or euro based on reserves). Although in theory their reserves are backed by traditional assets, transparency levels vary, and central banks have no direct supervisory control over how they are managed. Stablecoins are gradually becoming a way to hold money and make cross-border payments, bypassing traditional banks. Major players (Tether, USDC, others) manage sums comparable to the GDP of small countries, often operating in legal gray zones.

• Stablecoins actively compete with bank deposits for retail and corporate investor funds, especially in countries with unpredictable macroeconomic conditions
• A panic on the crypto market could cause an instantaneous outflow of traditional money from the banking sector
• Regulating them through standard banking channels is currently extremely difficult due to their distributed nature
• A sharp change in stablecoin demand could transmit a shock to the real economy through the dollar or euro

Escrivá insists that central banks cannot remain observers but must be more active in setting rules in this area, acting not just as regulators but as guarantors of financial stability. This requires new tools and authorities.

What needs revision

This is not about banning innovation, but about updating infrastructure for new realities. Comprehensive measures are needed:

• New stress tests for banks that explicitly account for AI scenarios and algorithmic failures
• Study of the interaction between traditional banks and the growing crypto ecosystem under extreme market conditions
• Update of international regulatory frameworks for stablecoins with clear requirements for reserves and reporting
• Cooperation among central banks to synchronize responses to AI shocks and avoid arbitrage between jurisdictions

The European Central Bank has already begun preparing framework documents in cooperation with other central banks, but Escrivá hints that the current pace lags behind the speed of fintech innovations. If they don't move faster, regulators may fall behind.

What this means

Central banks are gradually transitioning from a reactive position to a preventive one. Previously, they mainly reacted to crises. Now they recognize that AI and crypto are not marginal phenomena but parts of the financial system requiring new security frameworks. For markets, this could mean an acceleration of crypto sector regulation (which could be unfortunate for some startups) and possible slowdown of certain AI application adoption in finance until adequate safety levels are achieved.