6 Organizational Governance for AI‑Assisted Analysis

This chapter outlines acceptable use, accountability, and disclosure norms for generative AI in policy and health analytics.

6.1 Acceptable uses

• Scoping and question formulation
• Drafting and revising text
• Surfacing assumptions and gaps

6.2 Non‑acceptable uses

• Generating or validating evidence
• Making policy decisions
• Replacing expert review

6.3 Accountability

Human analysts remain fully responsible for AI‑assisted work.

6.4 Transparency

Disclosure should be proportionate and focused on interpretability and trust.