The Book on AI Doc Review

Overview

This book is your comprehensive, step by step guide to leveraging Artificial Intelligence in document review, covering technology, security, privacy, and workflows.It contextualizes AI within the evolution of Predictive Coding technology and legal workflows.

You’ll gain the expertise to confidently use AI for document review in eDiscovery, ensuring efficient and defensible processes aligned with legal requirements.

 “AI Doc Review” focuses solely on using AI to produce relevant documents and winning your case while maintaining unquestionable defensibility in court.

We once had AI review a large volume of data from Tim Kaine’s time as governor of Virginia, searching for documents related to the Virginia Tech shooting. A document that was returned as relevant said,

“I’m sorry about everything that has happened. I’ll see you at the memorial service tomorrow. God bless.”

So, I’m looking at this document, which I categorized as Not Relevant, and could not, for the life of me, figure out what the AI was thinking. I read the explanation, and it said,

“This document is dated 3 days after the Virginia Tech shooting. Given the memorial service and the context, it is likely they are referring to the Virginia Tech shooting.”

It blew my mind!

Find Answers In The Book

Isn’t TAR and Generative AI the Same?

TAR requires human trainers to classify documents as positive or negative examples, while Generative AI doesn't rely on predefined training examples, using natural language instructions instead, making it a faster and more efficient process.

As Attorneys, Can we Trust AI with our Data?

As with any technology, we must ensure that we are only using enterprise-grade tools built with security in mind. If used properly, there are no more risks with AI than any other technology.

How Does AI Doc Review Work?

While theory is great for some things, the only way to understand a process is to walk through each step. This book will walk through every step of the process.

Is AI Review Defensible?

Absolutely, but it must be done properly. This book will walk you through the process of running an AI review in a defensible manner.

About the Author

Jim Sullivan is the co-founder of eDiscovery AI. He is an attorney who has worked in eDiscovery since graduating from law school in 2007 and would never miss an opportunity to nerd out about technology.

Jim has consulted on thousands of Predictive Coding projects and frequently speaks at legal conferences about the use of technology in eDiscovery. He lives in Minnesota with his wife and daughter.