Industry Insight: Why the Future of eDiscovery is ‘All About AI Models’

“The key to the future is the past. What I mean by that is that the future is models – artificial intelligence models.”

That is just a taste of what NexLP’s CEO Jay Leib shared when he spoke as a featured panelist at ILTACON 2018 last month during a session that explored the future of eDiscovery.  This panel dove into industry leaders’ predictions on what will disrupt this market, and how organizations should be adapting. They also shared insights into how law firms can prepare themselves to stay competitive in the eDiscovery marketplace using available legal tech.

“There’s no bigger question we have here then the future of eDiscovery,” Leib told the crowd. He compared how we are talking about the future of legal tech to how we used to talk about something as simple as plastics. Referencing a scene from the movie “The Graduate,” where Dustin Hoffman’s character is told “there’s a lot of future in plastics.”

Of course, plastics aren’t quite as innovative as an industry as they once were. But there was a time where “plastics” was changing how entire industries were run. Innovation, in its basic sense, all starts from a simple concept. Today, technology has accelerated the rates at which organizations must adapt in the market.

“That’s what we’re going to be thinking about in the future…the simpler times. The future is AI,” Leib said, discussing about how years of manually collecting and analyzing data has prepared us for this much needed shift toward automation and augmentation of human behavior.

“In this room today, there is a treasure trove of what was previously sunk costs. Cases that were worked on before…contracts, agreements, etc. These are the stepping stones — the bootstrapping to models moving forward,” Leib said.

It’s been two decades since modern “electronic discovery” tactics were first being discussed. The problem, Leib argues, is that we’re “still using the same paradigm today of reviewing documents for the purposes of finding relevant documents, or agreeing for contracts or transactions.” But models aren’t set in stone, they are constantly adapting — and so should the industry.

“Models are to help decide what is relevant…what’s not relevant, what’s not relevant…what contracts should I sign or not sign. What mailboxes should I put on hold automatically...These will all be built up from the data of the past to be built out and build probabilities of the relevant documents in the future,” Leib said.

These models are exactly how eDiscovery teams help differentiate themselves, while creating new revenue streams for themselves and their clients.

“I believe with AI models, all professional service industries and companies are under threat — service firms, accounting firms, law firms, etc. I believe that the future is going to be an arms race. That law firms, accounting firms, consulting firms will be looking to build out models. Instead of having their advertisement done through people, what they’ll be doing is advertising their models,” Leib said.

“What will happen in this future that I’m predicting out is that new revenue streams will emerge. Law firms, accounting firms, consulting firms. etc., will start to take these models that they’ve built and be able to have new revenue streams within them. They’ll offer them up as subscription services through their clients. They’ll be able to take their expertise and put it at scale for their customers.”

That’s the future of eDiscovery in a nutshell, Leib said. As technology adapts, this is going to change how teams think about the role of technology in their everyday professional tasks. This starts by finding technology that is not only ubiquitous, but invisible.

“What you really want is to be invisible. The AI machine learning is behind the scenes. They are available to the end user. This is the consumerization of technology,” Leib said. “The same sort of UI/UX experience is this combination of of UX machine learning together is what’s really happening. That’s where we need to work together as an industry.”

The problem with eDiscovery processes is that they’ve been entirely too reliant on input from actual legal teams. This is an ineffective approach since it doesn’t account for one key component that AI models help humans overcome: Searching for things that you don’t actually know you need to look for. Many eDiscovery platforms were first launched at the same time the internet was, which left a gap in the marketplace.

“Today, everything is data. From email to chat messages, it’s all relevant. We live in a world of perfect evidence. The data needs to be automatically analyzed and gathered.”’


To catch the rest of Leib’s talk, as well as the remainder of the eDiscovery panel, catch the audio recording here.

Catch our recap of ILTA — and what you may have missed — from the conference here.