This repository supports the book Practical Legal Applications of Generative AI (AI LAW), an educational, governance-first guide for U.S. legal practice. It is designed to help practicing lawyers build a clear mental model of what generative AI can do today, where it fails, and how to use it safely under professional responsibility constraints.
- Release page: https://github.com/alexdibol/ai-law/releases/tag/book-ai-law-v01
- Direct download: https://github.com/alexdibol/ai-law/releases/download/book-ai-law-v01/AI_LAW.pdf
- Folder in this repo: https://github.com/alexdibol/ai-law/tree/main/notebooks
- Chapter 1: https://colab.research.google.com/github/alexdibol/ai-law/blob/main/notebooks/chapter_1.ipynb
- Chapter 2: https://colab.research.google.com/github/alexdibol/ai-law/blob/main/notebooks/chapter_2.ipynb
- Chapter 3: https://colab.research.google.com/github/alexdibol/ai-law/blob/main/notebooks/chapter_3.ipynb
- Chapter 4: https://colab.research.google.com/github/alexdibol/ai-law/blob/main/notebooks/chapter_4.ipynb
- Chapter 5: https://colab.research.google.com/github/alexdibol/ai-law/blob/main/notebooks/chapter_5.ipynb
- U.S.-based practicing lawyers and legal professionals
- Readers with minimal AI background who want practical workflows, not hype
- Teams building policies, templates, and controlled processes for AI-assisted legal work
This book treats generative AI not as legal authority, but as a tool whose usefulness depends on disciplined workflows, verification, and accountability. It presents a five-level maturity ladder that progresses from individual assistance to organization-grade operations:
- Level 1 — Chatbots: supervised drafting and client communication support
- Level 2 — Reasoners: structured issue spotting, IRAC-style analysis, argument mapping
- Level 3 — Agents: multi-step workflows under human approval gates (human-in-the-loop)
- Level 4 — Innovators: reusable assets such as playbooks, clause libraries, checklists, and adversarial tests
- Level 5 — Organizations: an end-to-end operating model where a team can routinely explain what was produced and how—redacted inputs, prompts/roles, risks flagged, verification tasks, and sign-off responsibility
The focus is professional defensibility. Each chapter emphasizes controls that make AI outputs safer to review and harder to misuse:
- Strict separation of generation vs. verification
- Explicit labeling of facts provided vs. assumptions vs. open questions
- Zero-tolerance posture toward invented authority (no fabricated citations, cases, statutes, or rules)
- Governance-native artifacts that make work replayable and auditable: run manifests, prompt logs, risk logs, verification checklists, deliverables bundles, and sign-off records
To keep learning concrete, every level is demonstrated using the same four mini-cases:
- Criminal
- Regulatory / Administrative
- International
- Teaching / Academia
Each chapter ends with a minimum standard checklist that can be adopted immediately in real practice settings.
This repository and book are educational materials. They are not legal advice. Human lawyer review is required for any reliance-bearing use. Confidentiality, privilege, competence, supervision, candor, and client communication obligations apply at every step.
Do not paste sensitive client information into external model prompts; use redaction and “minimum necessary” inputs by default.
notebooks/— Companion Colab notebooks (chapter_1 … chapter_5)docs/— Landing page for GitHub Pages- Releases — Versioned PDF book editions and optional bundles
This collection is provided for educational and research purposes only. It does not constitute investment advice, trading advice, or a recommendation to engage in any particular trading strategy or financial activity.
Market behavior is uncertain, and any application of the ideas, code, or methodologies presented here is undertaken entirely at the reader’s own risk.
Artificial intelligence tools may have been used to assist in aspects of editing, code generation, formatting, or drafting during the development of this collection.
However, conceptual design, methodological choices, system architecture, editorial judgment, integration, and final responsibility for the content, structure, and conclusions of the work remained under direct human control throughout.
The author assumes full responsibility for the coordination, design, and construction of the collection as a whole.
This project is released under the MIT License.
Alejandro Reynoso, Practical Legal Applications of Generative AI (AI LAW), companion notebooks repository, GitHub.