AI Audit

Full definition

Audits range from narrow bias audits (NYC Local Law 144 for hiring AI) to comprehensive conformity assessments (EU AI Act for high-risk systems) to system-of-management audits (ISO/IEC 42001). Typical scope includes: data lineage and consent, model card review, evaluation results, fairness metrics, drift monitoring, security testing (e.g., red-teaming), and incident logs. Audit is the verification mechanism that makes governance real.

Why it matters

Audit-readiness is now a procurement gate. Public-sector buyers, healthcare payers, and large financial institutions increasingly require evidence of AI audit before contract signature. Organizations that bake audit artifacts into the development process — model cards, eval reports, lineage logs — pass audits in days; others spend months scrambling.

Example

Under NYC Local Law 144, every employer using AI in hiring must commission an independent annual bias audit and publish summary results before using the tool — a model now spreading to other US states and to EU AI Act conformity assessments.

Related terms

• AI Governance
  AI governance is the framework of policies, roles, controls, and processes an organization uses to ensure its AI systems are lawful, safe, fair, accountable, and aligned with business intent — across the full lifecycle from problem framing to retirement.
• AI Bias
  AI bias is systematic, unfair difference in an AI system's outputs across demographic, geographic, or other groups — usually caused by biased training data, biased labels, or biased problem framing rather than the algorithm itself.
• EU AI Act
  The EU AI Act (Regulation (EU) 2024/1689) is the European Union's comprehensive, risk-tiered regulation of AI systems, the world's first horizontal AI law, with obligations phasing in from February 2025 and full general-purpose AI rules applying from August 2025.
• Explainability (XAI)
  Explainability (XAI) is the property of an AI system whose decisions can be understood by humans — through model-level documentation, prediction-level attributions, and counterfactual explanations — and a regulatory expectation in finance, healthcare, hiring, and other high-stakes domains.
• Algorithmic Accountability
  Algorithmic accountability is the principle that a specific person, role, or organization is identifiable and answerable for the design, deployment, outcomes, and harms of an automated decision system — and that the mechanisms to enforce that answerability exist.