# Build judgment before the consequential decision.

Draft learning guides for directors, committees and executives. These materials are educational and await editorial approval.

## The Board Director’s Guide to Responsible AI Governance

Begin with a system inventory and business purpose. Ask who is accountable, which decisions are reserved, how risk appetite is expressed and what evidence would change the board’s decision.

## The Responsible AI Oversight Checklist

For each material system: name the owner; identify affected people; classify risk; document authority; review fairness, privacy and security evidence; test escalation; assign remediation; schedule review.

## Governing Autonomous AI Agents

Specify action boundaries, financial and data limits, approval triggers and revocation. Test denied actions, retries and delegation. Require a receipt linking the decision, policy version and observed outcome.

## Responsible AI Evidence and Assurance

Separate source facts, participant attestations, calculations and independent conclusions. Ask about sampling, freshness, reviewer independence and scope. An evidence link alone does not establish assurance.

## Responsible AI Governance in the Agentic Enterprise

Review interconnected agents and third-party dependencies as a system. Stress-test authority propagation, correlated failures and business continuity. Give executives clear escalation responsibilities.

## Explore

- [Responsible AI Observatory](/responsible-ai)

- [Standards inform governance. Scope determines applicability.](/responsible-ai/standards)

- [Research that makes governance examinable.](/responsible-ai/research)

- [Make risk accountable.](/responsible-ai/risks)

- [Build judgment before the consequential decision.](/responsible-ai/board-education)

- [Make the method visible.](/responsible-ai/methodology)

- [Comparable evidence comes before comparison.](/responsible-ai/benchmarks)