Robots Are Your Friends: The Need for Artificial Intelligence in eDiscovery
I saw this headline on another legal blog a few days ago: AI may be your new Co-Worker! Which of course led to this week’s eDiscovery Blues™ cartoon. For years now, people have been talking about the Rise of the Robot Lawyers in the legal tech industry, but so far, Skynet still hasn’t happened.
In fact, a 2020 survey by Deloitte found that of 1300 CIOs who participated, 60% said Artificial Intelligence (AI) “would assist rather than replace workers.” And there’s no doubt we can use the help.
With the continued growth of data, human efforts alone won’t be enough to manage all of the information created by individuals, businesses, and government agencies. Doug Austin at eDiscoveryToday just published a great piece on Artificial Intelligence (AI) and its effect on eDiscovery.
He states that we can expect “463 exabytes (over 463 million terabytes!) of data to be created each day globally by 2025. Artificial Intelligence (AI) plays a big part of that reason for the mushrooming growth of personal data. It’s a big mess. Now, with data privacy laws strengthening, since AI helped get us into this mess, can it help get us out?”
In eDiscovery we’re already well aware of the various types of machine learning – clustering and predictive coding are the most prevalent – which legal teams can use to cull large datasets, identify PII, and locate responsive Electronically Stored Information (ESI). But there are still many in-house legal teams and law firms out there who try to apply more manual workflows to these situations without recognizing the value AI can bring.
It’s difficult to put an exact dollar-amount on how much savings a particular technology creates for legal teams, because every matter is unique, with infinite variables involving discovery parameters, data volume, and data types. But Stephen Goldstein, Director of Practice Support at Squire Patton Boggs, speaking at the 2019 Ipro Tech Show, gives a general breakdown of how using predictive coding adds value.
“If you start with 500,000 documents—and they’ve already been de-nisted and threaded, and you’ve gotten down to the core of the data you need to work with—a very good keyword approach would reduce that by about 65%, and you’d still have 175,000 documents. If that is sent to a managed review service, they could plow through them for a charge of around $180,000 for that kind of project, give or take. And that is assuming none of that work comes internal to lawyers who bill $400 an hour but is based on a $40 an hour charge.
“The predictive coding scenario is one that I’m very familiar with, and it’s one that we use. We would probably get an 85% reduction in the data, with fewer documents to review, at a far less cost. And we would have the benefit of that technology being in place to help with QC.”
This benchmark of saving a dollar for every document not sent for outside review – or as I heard one attorney say “a buck a doc” – is important to remember for both corporate in-house legal teams and law firms alike. But to do that with such large datasets, while protecting PII and meeting compliance requirements, you must leverage technology. No need to fear the robot overlords – say hello to your new AI assistant.
Learn more about Ipro’s AI & Advanced Analytics!