I’ve been meaning to cover this! Will ChatGPT replace human review? This article goes in depth to consider what it can and can’t do.
The article from Legaltech® News (What Will Ediscovery Lawyers Do After ChatGPT?, available here and written by John Tredennick and Dr. William Webber of Merlin Search Technologies) does a great job of analyzing the capabilities of ChatGPT and how they relate to eDiscovery capabilities. They begin by asking ChatGPT “How might ChatGPT Impact eDiscovery in the next ten years?” and part of its response was this:
“ChatGPT could be used to quickly identify relevant documents and extract key information from them, reducing the need for manual review and allowing legal teams to focus on higher-level tasks, such as strategy and analysis.”
ChatGPT did add that “it’s important to note that AI is not a replacement for human review, rather a tool to assist the process.”
So, will ChatGPT replace human review and make eDiscovery lawyers obsolete? As Tredennick and Webber state: “We doubt it, but this new generation of deep learning AI tools may ultimately play a role in how we analyze and review large quantities of documents, reducing review costs in the bargain.”
Tredennick and Webber proceeded to ask ChatGPT several questions, such as “What is required for a privilege log in ediscovery?”, how to write a certain Boolean query and how Enron violated US Federal Government accounting standards. They even provided a couple of documents from the Enron case and it successfully determined whether they were responsive to a specifically worded production request.
The results were less impressive when they conducted a more formalized test of documents from the Jeb Bush collection, using ChatGPT’s underlying GPT-3 engine, as it was “reluctant” to classify documents as relevant, whether they had been previously determined as highly relevant, relevant, irrelevant or unreviewed (presumed irrelevant as they hadn’t been identified by predictive models).
Tredennick and Webber also discuss several drawbacks to ChatGPT, including that it “isn’t always right”, which I discussed in my own analysis here. So, at least as it currently exists, the answer to the question “will ChatGPT replace human review?” is “no” (at least for now). Check out the terrific article with the very detailed analysis by Tredennick and Webber here!
So, what do you think? How do you think ChatGPT might impact eDiscovery in the next ten years? Please share any comments you might have or if you’d like to know more about a particular topic.
Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by my employer, my partners or my clients. eDiscovery Today is made available solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscovery Today should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.
Thanks for a thoughtful post on our article. We are doing further research on the role of ChatGPT in ediscovery and will publish that work soon
In the meantime, we are becoming increasingly doubtful that ChatGPT will take humans out of the picture. It may instead go down as the George Santos of ediscovery.