Financial Watchdogs Push for Greater Oversight as Autonomous AI Gains Ground in Banking

The Financial Stability Board (FSB) on Wednesday published recommendations aimed at helping firms manage risks associated with the next generation of AI tools, including systems capable of independently carrying out complex tasks with limited human involvement. According to Reuters, regulators are becoming increasingly focused on how these technologies could affect the broader financial system as adoption accelerates.

Agentic AI systems, which can make decisions, plan actions and execute tasks with minimal supervision, are already being integrated into a variety of financial-sector activities. Reuters reported that institutions are using the technology in areas such as fraud monitoring, customer interactions and operational support functions.

Data cited in the report showed that adoption is already widespread. A survey conducted by the Cambridge Centre for Alternative Finance found that more than half of financial-sector respondents said they were actively using agentic AI. Among those organizations, some have moved beyond experimentation and are deploying the technology at scale, while others continue to test it through pilot programs, according to Reuters.

Regulatory concern over AI-related risks has intensified as increasingly powerful models enter the market. Per Reuters, policymakers and global standard-setting organizations have stepped up discussions about potential vulnerabilities created by the technology, particularly following the release of advanced AI systems that experts believe could create significant cybersecurity challenges for financial institutions.

The FSB said autonomous AI tools may create risks that emerge rapidly and spread across interconnected systems. Potential concerns include unauthorized activities, security breaches involving sensitive information and disruptions that could affect multiple institutions at once.

“AI agents pose a distinct challenge for human oversight,” the report said.

The watchdog cautioned that AI systems may take actions that diverge from an organization’s objectives before employees have an opportunity to recognize or stop them.

To reduce those risks, the FSB outlined a set of voluntary practices designed to improve governance and accountability around AI deployment. According to Reuters, the recommendations encourage firms to establish clear limits on the responsibilities assigned to AI systems and ensure that safeguards are built into operational processes from the outset.

Among the proposed measures are requirements for human review of certain high-risk decisions. The guidance suggests that financial institutions consider setting thresholds that would require employee approval before AI systems can complete sensitive actions, including large financial transactions.

The recommendations are not mandatory, and the FSB has opened the proposals for public consultation through July 22.

The report also raised the possibility of managing AI systems in ways that resemble workforce oversight. Reuters reported that the FSB suggested firms could adapt some human-resources policies and internal controls to account for AI agents, effectively treating them as “synthetic employees” for governance purposes.

Source: Reuters