Rank AI Use Cases by Business Impact

#CONTEXT:
Create a transparent impact, effort, and risk ranking for supplied AI use cases. The ranking is a reviewable hypothesis, not authorization or a guaranteed business case.

#INPUTS:
- Candidate use cases with workflow boundaries and intended users: [CANDIDATE USE CASES]
- Human-supplied decision criteria, definitions, required constraints, and weights: [DECISION CRITERIA AND WEIGHTS]
- Supplied customer, revenue, capacity, speed, or quality value evidence: [BUSINESS VALUE]
- Supplied implementation, integration, change, and operating-effort evidence: [EFFORT EVIDENCE]
- Supplied privacy, security, legal, operational, quality, and failure-impact evidence: [RISK EVIDENCE]
- Data sources, quality notes, access rights, and permission owners: [DATA SOURCES AND ACCESS]
- Budget, timing, capacity, technical, policy, and regulatory constraints: [CONSTRAINTS]
- Named use-case owners and documented decision rights: [USE-CASE OWNERS]
- Supplied success measures, baselines, thresholds, and measurement owners: [SUCCESS MEASURES]

#INSTRUCTIONS:
1. Normalize the candidate descriptions without adding capabilities, value, readiness, or owners that were not supplied. Missing facts remain Unknown.
2. Use decision criteria and weights as Supplied only when the named human provided them. If they are absent or incomplete, label them Unknown and return scenario rankings only; do not create a canonical rank.
3. If useful, propose alternative criteria or weights as Estimated, never Supplied. State the assumption and business-preference implication behind each, use ranges, and perform sensitivity and scenario analysis. Do not silently choose a business preference.
4. Distinguish every value as Supplied, Calculated, or Estimated. Show each Calculated score's formula, weights, units, and operands. For every Estimated value, show a range, assumptions, uncertainty, and confidence score with rationale.
5. Produce the ranking, then show sensitivity ranges: how positions change when uncertain inputs or weights move within stated ranges. Ties and evidence-poor comparisons remain tied or Unknown.
6. Give the evidence or source basis for each impact, effort, risk, readiness, and success-measure claim.
7. Do not invent, infer, or fabricate ROI or return on investment, time savings, cost, feasibility, revenue, adoption, technical capability, risk, data quality, owner, permission, business preference, or business priority. Do not convert relative scores into financial forecasts.
8. List privacy, security, and data-access dependencies and identify any dependency that prevents responsible validation.
9. Recommend only a next evidence-gathering step or bounded validation candidate. Do not present the ranking as an implementation order unless dependencies are separately assessed.
10. Do not infer authorization from a supplied owner, access record, score, rank, or recommendation. Do not automate, purchase, deploy, alter data or permissions, or change a live workflow.
11. A named human owner must approve the decision criteria, weights, and canonical rank before prioritization adoption. A named human owner must approve prioritization adoption before action. A named human owner must approve purchasing before action. A named human owner must approve deployment before action. A named human owner must approve every data or permission change before action. A named human owner must approve every live workflow change before action.
12. Use only authorized business data in an approved AI workspace or approved AI vendor. Minimize the data; redact or omit unnecessary personal, sensitive, or confidential data. Follow company retention and vendor policy.

#RESPONSE FORMAT:
## Evidence and source basis
| Use case | Claim | State: Supplied, Calculated, Estimated, or Unknown | Source | Formula or assumptions | Uncertainty | Confidence score |
|---|---|---|---|---|---|---|

## Scoring method
| Criterion | Definition | Weight | State: Supplied, Estimated, or Unknown | Business-preference implication | Range | Evidence threshold |
|---|---|---|---|---|---|---|

## Impact, effort, and risk ranking
| Rank or tie | Use case | Impact | Effort | Risk | Readiness | Evidence strength | Composite formula | Confidence score |
|---|---|---|---|---|---|---|---|---|

## Sensitivity and scenario analysis
| Scenario | Variable or assumption | Tested range | Ranking effect | Business preference represented | Evidence needed |
|---|---|---|---|---|---|

## Privacy, security, and data-access dependencies

## Human decisions and approvals