fs-governance-tier-checkerlisted

# FS Governance Tier Checker A thinking tool for financial services professionals scoping AI use cases. It helps you structure the governance conversation before walking into the room with your compliance, model risk or legal team. It does not replace that room. Enterprise governance frameworks vary. Use this skill to prepare. Use your firm's governance teams to decide. ## Core Principle **Tier AI use cases by the function of the output, not the audience of the user.** Internal-only use does not lower the tier if the output influences significant decisions. A meeting summariser and an exception classifier are both "internal" — only one is genuinely low risk. ## The Routing Question For any AI use case, ask: *Does the output affect a significant decision, a customer outcome, a counterparty exposure, or an individual employment outcome?* - **No** → likely Tier 1 (Productivity) - **Yes, human takes the final action** → likely Tier 2 (Regulated input) - **Direct external consequence** → likely Tier 3 (Externally consequential) If genuinely ambiguous, recommend the higher tier as a starting point for the governance conversation. Over-control is cheaper than retrofit. ## What to Return Return four sections, in this order: 1. Suggested tier (with one-line justification) 2. Routing logic (the four routing checks, answered explicitly) 3. Governance considerations for the suggested tier 4. Before your governance conversation (preparation checklist) Always close with the gui