FSB Warns Frontier AI Models Could Increase Cyber and Financial Stability Risks

The Financial Stability Board (FSB) has warned that the rapid development of frontier artificial intelligence models could create new risks for the global financial system, with cyber threats emerging as one of the most immediate concerns.

In a letter to G20 finance ministers and central bank governors, FSB Chair and Bank of England Governor Andrew Bailey called on authorities to support the safe and responsible release and deployment of increasingly capable AI models.

The warning comes as financial markets are already facing several sources of vulnerability, including high asset valuations, growing leverage, concentrated market positions, and rising optimism around AI-related investments.

Frontier AI Adds a New Layer of Risk

Frontier AI models are becoming more autonomous and capable of solving complex problems. Their growing ability to operate with limited human intervention is opening the door to new applications across financial services.

At the same time, those capabilities could also be exploited.

The FSB identified cyber risk as a particular concern. More capable AI systems could potentially make cyberattacks more sophisticated, faster, and easier to scale. Financial institutions, which depend heavily on interconnected technology systems and third-party providers, could face greater exposure as these capabilities advance.

Bailey urged authorities to take appropriate measures to encourage responsible model development, release, and deployment.

For financial firms, the FSB is emphasizing preparation rather than simply prevention. Institutions should maintain strong response and recovery capabilities so they can continue operating if an AI-related cyber incident occurs.

The resilience of critical technology providers is also becoming increasingly important.

AI Optimism Meets Rising Market Leverage

The FSB’s concerns extend beyond cybersecurity.

Global financial markets are already showing signs of vulnerability. The organization pointed to elevated valuations, sovereign debt pressures, private credit vulnerabilities, and high expectations surrounding AI-related investments.

At the same time, the use of leverage in equity markets has increased.

The FSB warned that these factors can reinforce one another. Strong investor enthusiasm for AI can push valuations higher, while leverage can increase the impact of a sudden reversal.

If market sentiment changes sharply, highly leveraged positions could amplify losses and contribute to a disorderly correction.

Because financial markets are closely connected across borders, stress in one market could spread quickly to other jurisdictions.

Geopolitical and Inflation Pressures Add to the Challenge

The warning comes against a difficult macroeconomic backdrop.

According to the FSB, the conflict in the Middle East has added to energy-related inflation pressures. At the same time, concerns around sovereign debt and other areas of the financial system remain.

That means policymakers are dealing with several risks at once.

AI is introducing a new technological dimension, while traditional financial vulnerabilities have not disappeared.

This combination makes resilience particularly important for banks, asset managers, and other financial institutions.

Financial Institutions Need Stronger Resilience

The FSB is encouraging financial institutions to strengthen their ability to respond to technology-related disruptions.

That includes maintaining effective incident response and recovery plans, as well as assessing the resilience of important third-party technology and common service providers.

The message is particularly relevant as financial firms increasingly rely on cloud infrastructure, AI platforms, external data providers, and other technology partners.

A disruption at one major provider could potentially affect multiple institutions at the same time.

For financial companies adopting advanced AI, governance will therefore need to extend beyond the model itself. Firms will also need to understand how AI interacts with their wider technology infrastructure, cybersecurity systems, and operational processes.

Regulation Must Keep Pace With AI

The FSB said it is examining what steps it can take within its mandate and expertise to address the challenges associated with frontier AI.

The organization’s latest warning highlights a broader issue facing regulators worldwide: financial innovation is moving quickly, while supervisory frameworks must keep pace with new forms of risk.

The challenge is not necessarily to slow down AI adoption.

Instead, regulators and financial institutions need to ensure that the benefits of increasingly capable AI do not come at the expense of financial stability, cybersecurity, or operational resilience.

As AI becomes more deeply embedded in financial markets and institutions, the ability to manage its risks could become just as important as the technology’s ability to create value.

The Bigger Picture

The FSB’s warning arrives at a point when AI is moving from experimentation into critical financial infrastructure.

Banks and financial institutions are using AI for fraud detection, customer service, cybersecurity, risk assessment, trading, and compliance. More autonomous systems could eventually take on increasingly complex tasks.

That creates significant opportunities, but it also raises the stakes.

A powerful AI system can improve efficiency and decision-making. The same capabilities, if misused or poorly controlled, could increase cyber risk or accelerate market disruptions.

For regulators, the priority will be finding a balance between encouraging innovation and protecting the stability of the financial system.

For financial institutions, the message from the FSB is clear: AI readiness now needs to include risk readiness.

FSB Chair warns of risks arising from frontier Artificial Intelligence (AI) models