FSB Warns AI-Driven Cyber Risks Could Destabilize Global Markets
The G20 financial watchdog warns that frontier AI models may scale cyberattacks to a systemic level, threatening global stability.
The Financial Stability Board (FSB), the global financial watchdog for G20 economies, has identified AI-driven cyber risks as a significant threat to the stability of the global financial system. The organization warns that the rapid evolution of artificial intelligence could enable cyberattacks of unprecedented scale and sophistication.
According to the FSB, the primary concern lies in how frontier AI models can be leveraged by bad actors to automate the exploitation of vulnerabilities and enhance the effectiveness of phishing campaigns. This shift allows attackers to target financial institutions with a level of precision and volume that traditional security measures may struggle to contain.
The Dual-Use Dilemma
This threat emerges from a fundamental dual-use dilemma within the financial sector. The same AI technologies currently deployed by banks and regulators to detect fraud and secure networks are simultaneously available to cybercriminals. While AI enhances defensive capabilities, it also provides attackers with the tools to automate the discovery of software flaws and create highly convincing social engineering lures.
Systemic Implications
Regulators, including the Bank of England and the FSB, have warned that the highly interconnected nature of the global financial system creates a high risk of contagion. Because financial institutions operate across borders through shared infrastructure and payment rails, a major cyber-disruption fueled by AI in one jurisdiction could rapidly spread to others. Such an event could lead to widespread operational failures and a sudden loss of confidence in banking systems, potentially destabilizing global markets.
The Path Forward
As frontier AI models continue to advance, regulators are focusing on the systemic nature of these risks rather than treating them as isolated IT failures. The focus remains on how to build resilience against AI-powered threats that can evolve faster than human-led defense strategies. It remains to be seen how international regulatory frameworks will adapt to standardize AI security protocols across the G20 to prevent a coordinated systemic collapse.