Financial regulators urged to strengthen AI oversight

Financial regulators worldwide must strengthen their oversight of artificial intelligence and improve their ability to identify and contain emerging risks before its adoption outpaces regulatory safeguards, financial experts said.

They called for stronger risk management, greater technological expertise and closer international cooperation to address AI-related threats to financial stability.

Sheila Bair, a member of the International Advisory Council of China's National Financial Regulatory Administration, has urged financial institutions to take a cautious approach to adopting artificial intelligence until regulators develop a better understanding of its risks and how to contain them.

AI is equipping banks with more powerful tools to defend against cyberattacks, but it is also giving hackers new capabilities, Bair said on the sidelines of the Annual Conference of Financial Street Forum 2026, which concluded in Beijing on Thursday.

Other emerging risks include a lack of transparency and algorithmic bias, particularly in credit decisions, AI-enabled fraud, and concentration risk arising from reliance on a small number of AI providers, she said.

"It is absolutely essential that any use of AI be accompanied with clear governance, accountability and risk management," Bair, former chair of the US Federal Deposit Insurance Corporation, said.

However, she noted that financial and non-financial institutions worldwide are racing to reap AI's perceived benefits, often placing risk management second to the pursuit of commercial gains.

Bair also stressed the urgent need for information sharing protocols to help identify and counter malicious actors' use of AI, warning of its potential to disrupt financial infrastructure.

Andrew Sheng, another member of the NFRA's International Advisory Council and one of the Hong Kong SAR Chief Executive's Council of Advisers, said regulators must keep pace with technological advances and strengthen their ability to anticipate emerging risks.

Regulatory failures could occur if authorities lack a thorough understanding of the architectures, risks and development trajectories of large language models, distributed ledger technology and privacy-enhancing computing, Sheng said.

He urged regulators to strengthen employees' expertise and develop, refine and deploy intelligent, real-time supervisory tools capable of detecting algorithmic opacity, market anomalies and hidden connections.

Sheng warned that poor data integrity and inadequate governance could lead to flawed AI models that generate inaccurate outputs, or "hallucinations", resulting in misleading analyses or signals.

Regulators must clearly define data ownership, access rights and security boundaries, while balancing the protection of individual privacy and national security with the need to unlock the economic value of data, he said.

Sheng further warned that geopolitical tensions could lead to increasingly divergent AI standards and development models. Regulators must therefore strengthen international cooperation to address emerging differences in standards, protocols and algorithms.

Enhanced bilateral and multilateral information-sharing, together with coordinated crisis-response protocols, is essential to containing risk transmission across jurisdictions, he said.

The 2026 meeting of the International Advisory Council of the National Financial Regulatory Administration was held in Beijing from Monday to Tuesday.

The meeting called on financial regulators across countries to strengthen policy coordination and cross-border regulatory cooperation to better prevent and respond to cross-border risks amid rising uncertainty in the international environment. It also urged financial institutions to guard against emerging financial risks while leveraging new technologies, including AI, to improve efficiency.

In addition, the meeting stressed the need to further improve mechanisms for early warning and response to financial risks and to firmly uphold the bottom line of preventing systemic financial risks.