AI Literacy Curriculum Architecture

Purpose

The AI Literacy Curriculum is the canonical training framework for building AI competency across all legal department roles. It provides five role-specific learning tracks with structured modules addressing Defensible AI principles, professional responsibility, Risk Taxonomy 2026 awareness, AI BoM governance, and Agentic Tier supervision. Completion of foundational tracks is a prerequisite for participation in any AI pilot under USE-02.

Operating cadence: Continuous — foundational rollout in Band 1; role-specific deepening in Band 2; enterprise-wide integration in Band 3.

Owner: Legal Operations, HR/Learning and Development, STR-07 AI Task Force.

When to Use This Module

• Before launching any AI pilot — USE-02 Phase 1 requires evidence that pilot participants have completed the relevant track
• When onboarding new legal staff in an AI-active department
• When the organisation advances to a new maturity band
• When the AI BoM is updated with a new approved tool — affected role groups require tool-specific training
• When the Risk Taxonomy 2026 classification framework is updated

AI Bill of Materials — Curriculum Integration Requirement (Metric 0)

The AI Literacy Curriculum must be connected to the organisation’s live AI BoM:

| AI BoM Curriculum Requirement | Status |

|—|—|

| Curriculum includes a module on AI BoM registration and approved tool selection | Confirm |

| All five tracks include instruction on how to check the AI BoM before using any AI tool | Confirm |

| Track 5 (IT/Security) includes AI BoM maintenance and Shadow AI detection procedures | Confirm |

Shadow AI prevention: The AI BoM training module is the primary mechanism for reducing Class 6 Shadow AI risk. Staff who understand the AI BoM approval process are significantly less likely to use unapproved tools. Completion rates for this module must be tracked and reported to STR-07.

Programme Objectives

1. Ensure all legal professionals understand Defensible AI principles and their professional responsibility obligations under ABA Formal Opinion 512
2. Equip Legal Leadership to evaluate, approve, and govern AI deployments using the Legal AI Blueprint framework
3. Provide Practicing Lawyers with competency in AI-assisted legal work, output verification, and Risk Taxonomy 2026 awareness
4. Train Legal Operations to manage AI tools, track ROAI (Return on AI Investment), and administer the AI BoM
5. Prepare Support Staff to use approved AI tools within defined scope limitations
6. Enable IT/Security to administer AI tool governance, detect Shadow AI, and maintain AI BoM integrity

Track 1: Legal Leadership

Audiences: General Counsel, Practice Group Heads, C-Suite

Duration: 8 hours (4 modules x 2 hours)

Module 1.1 — Legal AI Strategy and Defensible AI Principles

• Defensible AI framework: what makes an AI deployment defensible
• Legal AI OS Blueprint structure: 8 pillars, governance model, maturity bands
• STR-07 AI Task Force: role, authority, and approval gates
• Board and client reporting obligations for AI deployments
• Risk Taxonomy 2026: executive overview of 9 canonical risk classes

Module 1.2 — Governance, Risk, and Professional Responsibility

• ABA Formal Opinion 512: supervisory obligations for AI
• ABA Rules 1.6, 1.1, 5.3, 3.1, 1.5 mapped to AI deployment scenarios
• EU AI Act: high-risk AI system obligations for law firms and corporate legal departments
• Defensibility Posture Statement (DPS): what it is and how leadership evidences it
• GOV-01 Annual AI Governance Review: leadership participation requirements

Module 1.3 — ROAI and Value Measurement

• Return on AI Investment (ROAI): the correct framework for AI value measurement
• USE-05 Baseline Metrics Capture: why baseline measurement precedes every AI deployment
• USE-04 Legal AI ROAI Dashboard: reading and interpreting quarterly reports
• STR-08 ROAI Matrix: how leadership uses the quadrant model for portfolio decisions
• Agentic Tier considerations: elevated governance obligations for autonomous AI

Module 1.4 — AI Portfolio Strategy and Maturity Advancement

• USE-01 Use Case Prioritisation Matrix: how leadership approves the AI portfolio
• Maturity band advancement gates: evidence required to move from Foundational to Integrated
• AI BoM governance: leadership accountability for AI inventory management
• Shadow AI response: what to do when Class 6 risk is identified
• DPS reporting: connecting AI governance activity to external defensibility evidence

Track 2: Practicing Lawyers

Audiences: Associates, Partners, Counsel, In-House Attorneys

Duration: 12 hours (6 modules x 2 hours)

Module 2.1 — AI Fundamentals for Legal Practice

• How large language models work: capabilities and limitations
• AI legal databases and research tools: appropriate use in legal practice
• Prompt engineering for legal tasks: structured output, citation verification
• AI output characteristics: hallucination, bias, and confidence scoring

Module 2.2 — Professional Responsibility and AI

• ABA Formal Opinion 512: competence and supervision obligations
• Supervision obligations for AI-assisted work product: Rule 5.3 applied
• Confidentiality in AI: what data can and cannot be input to AI tools
• Fee disclosure: when AI assistance requires client disclosure under Rule 1.5

Module 2.3 — Risk Taxonomy 2026 Awareness for Attorneys

| Class | Attorney Scenario |

|—|—|

| Class 1: Hallucination and accuracy | AI-generated case citations that do not exist — verification requirements |

| Class 2: Privilege and confidentiality | Client data entered into unapproved AI tools — Rule 1.6 breach scenarios |

| Class 3: Bias and fairness | AI contract review tools with systematic bias against certain clause types |

| Class 4: Privacy and data protection | GDPR/CCPA data in AI tools — cross-border transfer risks |

| Class 5: Supply chain dependency | Vendor lock-in; what happens when an approved AI tool is discontinued |

| Class 6: Shadow AI | Using non-approved tools; how to check the AI BoM |

| Class 7: Regulatory compliance drift | AI tools that become non-compliant as regulations change |

| Class 8: IP and licensing | Who owns AI-generated work product; model training on client work |

| Class 9: Operational resilience | AI tool outages; maintaining manual backup capability |

Module 2.4 — AI Tool Proficiency (Role-Specific)

• Using approved AI research tools (AI BoM-registered only)
• Document drafting with AI: review obligations and output verification checklist
• Contract analysis tools: setting scope, reviewing AI-identified issues
• AI-assisted due diligence: scope limitations and attorney sign-off requirements

Module 2.5 — Agentic AI Supervision

Required for attorneys using any Agentic Tier tools:

• What is an Agentic Tier AI system and how does it differ from standard AI
• Attorney supervision obligations for autonomous AI actions: Rule 5.3 in agentic contexts
• Kill-switch procedures: when and how to halt autonomous AI workflows
• Audit trail review: attorney obligations to review autonomous action logs
• Escalation triggers: when autonomous AI output requires attorney review before use

Module 2.6 — Quality Assurance and Output Verification

• AI output verification checklist: mandatory review steps before any AI-assisted work product is filed, sent, or relied upon
• Citation verification: process for confirming AI-provided legal citations
• Hallucination detection: practical techniques for identifying plausible-but-false AI output
• Client communication: when to disclose AI assistance to clients

Track 3: Legal Operations

Audiences: Legal Operations Managers, Project Managers, Process Analysts

Duration: 10 hours (5 modules x 2 hours)

Module 3.1 — AI BoM Administration and Governance

• AI Bill of Materials: structure, purpose, and maintenance responsibilities
• Adding tools to the AI BoM: approval workflow, required fields, Agentic Tier designation
• Decommissioning tools: exit procedures, data deletion confirmation
• Shadow AI detection: using AI BoM gaps to identify unapproved tool usage
• Reporting AI BoM status to STR-07 AI Task Force

Module 3.2 — USE-05 Baseline Metrics Capture

• Why baseline measurement precedes every AI deployment
• USE-05 framework: four measurement dimensions (operational, quality, financial, user experience)
• Data collection methodology: conducting baseline surveys, extracting operational data
• Shadow AI baseline: anonymous survey methodology for capturing existing unapproved tool usage
• DPS zero-state documentation: establishing the Defensibility baseline

Module 3.3 — ROAI Tracking and the USE-04 Dashboard

• USE-04 Legal AI ROAI Dashboard: administering monthly data collection
• ROAI calculation methodology per STR-08 ROAI Matrix
• Connecting USE-05 baseline to USE-04 post-AI comparison series
• Reporting to General Counsel and STR-07: quarterly ROAI update format
• When ROAI falls below target: escalation procedure and STR-08 quadrant re-evaluation

Module 3.4 — AI Pilot Administration (USE-02)

• USE-02 Phase 1 pre-pilot requirements: baseline complete, training complete, AI BoM registered
• Pilot scoping and boundary documentation
• Data collection during pilot: USE-04 Dashboard configuration
• Phase 2 gate: what Legal Operations must confirm before pilot expansion
• Pilot-to-deployment transition: governance checklist

Module 3.5 — Risk Taxonomy 2026 for Legal Operations

• Class 6 (Shadow AI): detection, reporting, and remediation procedures
• Class 9 (Operational resilience): SLA monitoring and escalation
• Class 5 (Supply chain dependency): vendor health monitoring in AI BoM
• GOV-03 Risk Register: how Legal Operations flags and tracks operational AI risks
• GOV-05 Incident Response: Legal Operations role in AI incident management