Extract from Esther Birnbaum’s article “The Future Is Now: The Case for Adoption of Generative AI Document Review in E-Discovery”
Recently, I was asked why there is so much resistance to adopting generative AI (GenAI) for e-discovery document review. My answer was simple: It’s because many have yet to see it in action.
The legal industry is no stranger to caution, especially regarding new technologies. Concerns about accuracy, transparency, bias, and even the dreaded AI “hallucinations” have dominated discussions. While these are legitimate considerations, they often lead to a cycle of hypothesizing rather than action. It is time for the legal field to move beyond hypothesizing and start testing and using GenAI in real-world e-discovery scenarios.
The Proven Advantages of GenAI for First-Level Document Review
The primary goal of a first-level document review is to evaluate a dataset to identify documents pertinent to a case or specific issues. Utilizing a GenAI tool that analyzes and codes each document or issue for relevance can produce results that surpass those of a human review.
The benefits of GenAI for e-discovery are not merely theoretical—they are evidenced by real-world results. Testing has shown that GenAI delivers significantly higher precision and recall in document coding compared to human reviewers.
