AI in Finance: Global Watchdog Urges Tighter Controls

2026-06-10
AI in Finance: Global Watchdog Urges Tighter Controls

London – A global financial watchdog is calling for stricter controls on increasingly autonomous artificial intelligence (AI) systems being used in the finance sector. The Financial Stability Board (FSB) issued a report on Wednesday, highlighting concerns that these “agentic AI” systems – those capable of making decisions with limited human oversight – could significantly amplify risks to the financial system as their adoption continues to grow.

The FSB’s report specifically addresses the rising use of AI in areas such as trading, risk management, and customer service. While acknowledging the potential benefits of AI, including increased efficiency and innovation, the report stresses the need for proactive measures to manage the associated risks. The watchdog underscores the potential for AI to exacerbate existing vulnerabilities and create new ones, potentially leading to instability within the financial system.

The report “strongly” encourages boards of financial institutions to consider implementing safeguards to mitigate these risks. These safeguards could include enhanced monitoring of AI systems, robust testing and validation procedures, and clear lines of accountability for AI-driven decisions. The FSB also suggests exploring the development of industry-wide standards and best practices for the responsible use of AI in finance.

The call for tighter controls comes as AI technology rapidly evolves and becomes more integrated into financial operations. The FSB’s report reflects a growing global awareness of the potential risks associated with AI and a desire to ensure that the technology is used responsibly and in a way that supports financial stability. The report's findings are expected to influence regulatory frameworks and supervisory practices in countries around the world, including New Zealand, as financial institutions continue to explore and deploy AI solutions.

Recommendations
Recommendations