NIST AI RMF
APPLICATION
DATA
GOVERNANCE
INFRASTRUCTURE
Accountability structures
Description
Roles, responsibilities, and lines of communication related to AI risk management are established with clear accountability. Every AI system must have a named human accountable for its behavior in production.
Suggested Actions
1
Define AI governance roles with explicit authority: AI Ethics Officer (policy interpretation), Model Risk Manager (validation oversight), Data Steward (data quality and privacy), AI System Owner (business accountability for outcomes)2
Establish an AI Governance Committee or Board with executive representation, meeting at least quarterly, with authority to halt AI deployments that exceed risk appetite3
Document accountability chains from individual contributors to executive leadership — when an AI system causes harm, there must be zero ambiguity about who is responsible4
Create RACI matrices for every AI lifecycle activity: who is Responsible for execution, Accountable for outcomes, Consulted for input, and Informed of decisions5
Ensure accountability structures cover third-party AI systems — outsourcing the technology does not outsource the responsibility6
Integrate AI accountability into existing corporate governance structures (board risk committees, audit committees) rather than creating parallel governance silos