ICBA: Guidance on AI agent systems must reflect the realities of community banking

ICBA recommended to the National Institute of Standards and Technology that any future guidance on artificial intelligence agents reflect the realities of community banking.

Details: In a letter responding to NIST’s request for information on security considerations for AI agent systems, ICBA said:

• Agent systems should be treated as a higher-risk category of AI because they can take actions—not just generate text—that impact sensitive data, customer communications, and business operations.

• NIST should encourage strong identity orchestration to support a “Know Your Agent” approach so institutions can identify which agent is acting, what it is authorized to do, and who approved that authority.

• Guidance must emphasize "constrained deployments by default" in high-risk environments, clear accountability across the vendor ecosystem, and auditability that enables banks to investigate incidents, protect customers, and demonstrate appropriate oversight.

AI in the Spotlight:

• The Treasury Department last month released two resources to guide AI use in the financial sector.

• The Labor Department recently published a framework to guide nationwide AI efforts across workforce and education systems.

• President Trump in December proposed removing state-level AI laws and installing a national standard.

• Recent Independent Banker articles detail ways that community banks can use AI to fight fraud and tips on developing AI policies.