Use Case Prioritization Matrix

Purpose and Scope

This Module defines the Use Case Prioritization Matrix used by legal departments and law firms to select, rank, and periodically refresh AI implementation opportunities. It is the upstream input to USE-02 (Pilot Program Design) and STR-05 (Business Case and Cost of Inaction), and a key evidence source for GOV-01 / GOV-02.

It is designed for:

• Corporate legal departments (in-house)
• Law firms (all sizes)
• Hybrid or shared-services legal models

The goal is to focus limited resources on AI initiatives that can demonstrate measurable Return on AI Investment (ROAI) within 90–180 days while maintaining Defensible AI standards.

---

Section 1: Five-Dimensional Scoring Framework

Each candidate AI use case is scored 1–5 on five dimensions. Default weights can be customised to reflect organisational priorities.

1. Business Impact (Default Weight: 30%)

Definition: Quantifiable business value including cost reduction, efficiency gains, risk mitigation, and client service improvement.

Scoring scale:

• 5 – Transformational: >30% cost savings or major efficiency breakthrough
• 4 – Substantial: 20–30% improvement in key metrics
• 3 – Moderate: 10–20% improvement with clear value
• 2 – Minor: 5–10% improvement, still meaningful
• 1 – Minimal: <5% improvement, limited business case

Evaluation criteria:

• Time savings for legal professionals (hours per week)
• Cost reduction potential (annual dollars)
• Risk mitigation value (regulatory, litigation, operational)
• Client service enhancement (speed, accuracy, satisfaction)
• Process efficiency improvements (cycle time, throughput)

2. Feasibility (Default Weight: 25%)

Definition: Organisational readiness including data, technical fit, integration complexity, and user adoption likelihood.

Scoring scale:

• 5 – Excellent: high-quality data, simple integration, enthusiastic user base
• 4 – Good: adequate data, moderate complexity, willing users
• 3 – Fair: some data issues, standard complexity, mixed user readiness
• 2 – Challenging: limited data, complex integration, user resistance
• 1 – Difficult: poor data, major technical barriers, strong resistance

Evaluation criteria:

• Data quality and availability (completeness, accuracy, accessibility)
• System integration requirements (APIs, complexity, vendor constraints)
• User readiness and change management needs
• AI Bill of Materials (AI BoM) registration feasibility for candidate vendor
• Vendor ecosystem maturity and support quality

3. Speed to Value (Default Weight: 20%)

Definition: Time required to implement and demonstrate measurable ROAI.

Scoring scale:

• 5 – Immediate: 30–60 days to demonstrate clear value
• 4 – Fast: 60–90 days to show meaningful results
• 3 – Standard: 90–120 days for value demonstration
• 2 – Slow: 120–180 days to prove business case
• 1 – Extended: >180 days for ROAI demonstration

4. Risk Level – Risk Taxonomy 2026 (Default Weight: 15%)

Definition: AI-specific risk across the nine Risk Taxonomy 2026 classes. Lower risk exposure yields a higher score (5 = very low risk; 1 = very high risk requiring enhanced governance).

Risk classes:

1. Hallucination and accuracy
2. Privilege and confidentiality
3. Bias and fairness
4. Privacy and data protection
5. Supply chain and vendor dependency
6. Shadow AI and policy circumvention
7. Regulatory compliance drift
8. IP and licensing exposure
9. Operational resilience

Governance mapping:

• Score 4–5 per class: standard GOV-02 AI Use Policy controls
• Score 2–3 per class: enhanced monitoring; quarterly GOV-03 Risk Register review
• Score 1 per class: Agentic Tier controls; real-time monitoring; GOV-05 Incident Response plan required

High Agentic Tier use cases (autonomous multi-step execution without per-step human approval) automatically require enhanced governance under GOV-01 and explicit AI BoM entries.

5. Scalability Potential (Default Weight: 10%)

Definition: Ability to expand the AI solution across practice areas, matter types, and organisational units.

Scoring scale:

• 5 – Enterprise-wide: applicable across entire organisation
• 4 – Multi-area: scalable across several practice areas
• 3 – Practice-focused: expandable within a practice area
• 2 – Limited: narrow scalability potential
• 1 – Single-use: difficult to expand beyond pilot