Mechanisms that allow affected individuals to challenge or seek remedy for AI-driven decisions that impact their rights or interests.
Institutional processes - appeal workflows, human-review channels, automated explainers - that enable users to dispute AI outcomes (e.g., loan denials, content takedowns). Recourse governance defines SLA for appeal resolution, required transparency in outcome explanations, and metrics on recourse effectiveness (e.g., percentage of overturned decisions), ensuring organizations respect affected parties’ rights and improve AI accountability.
A credit bureau’s AI-based scoring portal provides applicants with an explanation of why they were denied and a “Request Review” button. Within five business days, a human underwriter reevaluates the case, offers corrections if warranted, and updates the applicant - meeting the organization’s recourse-policy commitments.
