AI-Assisted Review for Litigation Support: Workflows, Validation, and Defensibility

Course Overview

Generative AI review is now built into the major review platforms, and one of the first federal courts to rule on it treated it the way courts have treated technology-assisted review since 2012. That decision, Schulte v. LinkedIn (N.D. Cal. 2026), and a run of 2026 orders on AI prompts, AI chat logs, and protective orders that bar open AI tools have moved a new set of tasks onto the litigation support desk. Someone has to confirm the tool is approved for discovery data, build the review set, write and test the review instructions, measure the results, keep the record, and explain all of it to the lawyer who signs the discovery responses.

This course teaches that work step by step. You will learn what the case law and the federal and Florida rules expect, how to vet an AI tool before any data is loaded, how to pre-cull and prepare data, how to turn requests for production into review instructions and refine them against a lawyer-coded sample, how to run privilege screens, and how to compute recall, precision, and elusion from validation samples. The final section covers logging, preserving AI-generated content as ESI, and writing a methodology memo counsel can rely on. The course is vendor neutral and uses a fictional matter throughout. It is designed for litigation support professionals and is not offered for CLE credit.

What You Will Learn

• Explain how courts have treated technology-assisted review since 2012 and why the first generative AI review decision applied the same framework.
• Identify the obligations that ESI protocols, protective orders, and the federal and Florida rules place on the litigation support team.
• Evaluate whether an AI tool is open or closed and decide whether discovery data may be loaded into it.
• Apply search-term pre-culling and data preparation steps that keep an AI review set proportionate and complete.
• Construct review instructions from requests for production using the Five-Element Prompt and refine them against a lawyer-coded sample.
• Calculate recall, precision, and elusion from validation samples and interpret the results for supervising attorneys.
• Implement logging and preservation practices for prompts, model runs, and AI-generated content treated as ESI.
• Prepare a methodology memo that documents an AI-assisted review well enough for counsel to defend it.

Who Should Take This Course

• Litigation support analysts and managers at law firms
• eDiscovery project managers and consultants at service providers
• Corporate legal operations and discovery staff
• Senior paralegals who run or supervise document review

Course Outline

1. Section 1: The Legal Ground Rules
   • Lesson 1.1: From TAR to Generative AI Review
   • Lesson 1.2: The Rules Behind the Work
   • Lesson 1.3: Protective Orders, Privilege, and AI Artifacts
   • Quiz 1: The Legal Ground Rules
2. Section 2: Preparing Data for AI Review
   • Lesson 2.1: Vetting AI Tools Before Data Touches Them
   • Lesson 2.2: Pre-Culling with Search Terms
   • Lesson 2.3: Data That AI Review Handles Poorly
   • Quiz 2: Preparing Data for AI Review
3. Section 3: Running Generative AI Review
   • Lesson 3.1: Turning Requests for Production into Review Instructions
   • Lesson 3.2: Testing and Refining Against a Lawyer-Coded Sample
   • Lesson 3.3: Privilege Screening and Privilege Logs
   • Quiz 3: Running Generative AI Review
4. Section 4: Validation and Documentation
   • Lesson 4.1: Validation Sampling and the Numbers Counsel Needs
   • Lesson 4.2: Logging Prompts, Models, and Runs
   • Lesson 4.3: AI Content as ESI
   • Lesson 4.4: The Methodology Memo
   • Quiz 4: Validation and Documentation
5. LGTLS201 Final Exam (20 questions, 80 percent to pass)

Course Details

• Level: Intermediate
• Instruction: 71 narrated minutes (1 hour 11 minutes), plus quizzes and a final exam
• Prerequisites: Working familiarity with a document review platform and basic eDiscovery terms (custodian, processing, production)
• Format: Self-paced online course with narrated lessons, interactive exercises, section quizzes, and a graded final exam

Materials Included

• Student handbook (PDF and Word)
• Quick reference card: the AI review defensibility checklist and sampling formulas (PDF)

Continuing Education

This course is designed for litigation support and eDiscovery professionals. It is not designed or offered for CLE credit. A certificate of completion documents instructional time for professional development records.

How to Use This Course

• Complete the sections in order; each builds on the last.
• Work through every interactive exercise before taking the section quiz.
• Pass each quiz and the final exam with a score of 80 percent or higher to earn your certificate of completion.
• Keep the student handbook and quick reference card for use in practice.

Current as of November 2026. Product names, vendor policies, rules, and court decisions discussed in this course change; the course is reviewed on an annual cycle.