Global Financial Regulator Calls for Enhanced Human Oversight of AI Agents

The Financial Stability Board (FSB), an international body that sets global standards for financial regulation, has called on banks and other corporations to enhance human oversight of artificial intelligence (AI) agents. This requirement is included in the final guidance on AI risk management.

The FSB's guidance comes as regulators worldwide are increasing efforts to mitigate AI-related risks, including threats posed by AI-driven cyberattacks. A previous draft released in June had strongly advised financial institutions' boards to consider the recommendations, warning that AI agents could autonomously plan and execute tasks with limited human supervision, presenting unique challenges.

Based on feedback received and considering the increasing adoption of AI agents by financial institutions, along with the technology's inherent risks and control requirements, the final guidance strengthens oversight recommendations. It emphasizes that enhanced human supervision is crucial for higher-risk AI applications. Companies must also define, based on risk levels, when AI agents require human approval before proceeding with tasks.

The FSB also removed a previous suggestion that financial institutions should treat AI as a "virtual employee." Instead, it stresses that firms must recognize that AI agents do not think like humans and may interpret preset objectives in unexpected ways.