On the second day of Relativity Fest 2022, four AI and e-discovery experts gathered on stage, talking amongst themselves with clear rapport. These individuals—Omar Haroun, Relativity’s head of AI strategy and cofounder of TextIQ; Josh Kreamer, director of e-discovery at AstraZeneca; Laura Kibbe, assistant general counsel at IQVIA; and Bobby Malhotra, of counsel and co-leader of the Information Governance and e-Discovery Practice at Munger, Tolles and Olson LLP—were no strangers to one another. As AI innovators in the e-discovery industry, they participate in discussions together quite often, shedding light on the latest and greatest AI trends in legal technology.
For Fest attendees, the group decided to tackle new use cases for artificial intelligence, detailing the ways AI can be utilized outside of the traditional technology-assisted review workflows most familiar to the e-discovery industry. With Omar guiding the discussion, Josh, Lauren, and Bobby shared the exciting ways they solve complex, and at times daunting, legal challenges.
Intrigued? Read on to learn more about how these bright minds are leading the charge and leveraging AI to solve the most pressing data problems at their organizations.
Examining Confidential Data in an M&A Deal
Bobby Malhotra outlined the challenge Munger Tolles and Olson, LLP faced when they represented a client selling off a small division of their company. Many of the documents relevant to the transaction were intertwined with confidential information on other products and business units. There were also numerous documents where the client had confidentiality agreements with other third parties and, therefore, the documents could not be transferred over to the buyer.
“We needed to figure out a way to separate the confidential pieces of information and segregate the documents in a defensible manner,” described Bobby. “And we only had two months to get the deal done.” The team knew a traditional workflow of keyword searching and review would not be fast enough. Instead, they decided to develop a custom AI classification model for the case. The model was able to segregate the data defensibly and at a fraction of the cost.
While many legal professionals are familiar with confidentiality use cases and M&A transactions, few are currently using AI to assist. But Bobby was clear that, given the time constraints and the complex data components of the deal, AI was essential to a successful outcome: “Without the use of the classification model, we would not have been able to meet the deadline. The deal would have fallen through.”
Assessing Data Breach Exposure
In an environment of exponentially growing data and ever-present cybersecurity risks, a data breach is a top-of-mind concern. Laura Kibbe recalled when her company needed to proactively respond to a potential incident. As general counsel, she was responsible for assessing the company’s exposure and determining the best path forward.
The outside law firm she was working with wanted to review every document manually. “I said, ‘no way, we are using tools,’” said Laura. “We will defend the technology and show you how it works.”
Laura and her team recognized that they could reduce the amount of information they needed to review by using AI to set aside documents that did not contain personally identifiable information (PII) or protected health information (PHI).
“We wound up eliminating 90 percent of the material because it didn’t have PII or PHI,” recounted Laura. “We had the law firm validate those findings with a statistically significant sample. Then we agreed to review the final 10 percent.”
Overall, Laura and her team utilized AI to reduce the review population by 1.2 million documents, allowing the company to efficiently determine their data was not exposed and quickly return to business as usual.
Using AI for Data Governance
Toward the end of the session, an audience member broached the topic of data governance: “How can data governance and compliance teams best utilize AI?” he asked. “Is there a way that you can be confident you are keeping what you need, but the rest you can throw to the woodchipper?”
“There is an incredible opportunity related to information or data governance,” replied Omar. “Often there isn’t a whole lot of urgency to classify data until a litigation or breach presents itself. But if you think about ways to apply learnings from those scenarios to the future, that is exciting.”
Bobby strongly agreed, highlighting the importance of using AI for information governance “in real time. If you can classify a document and figure out what you do and don’t need before litigation, then your costs are less in the future. For example, certain documents wouldn’t need to be reviewed because they are already classified. In the information governance use case, it’s astonishing how the use of AI can impact things down the line.”
Thinking Outside the Box
To close, each panelist championed the importance of innovation and proactivity.
“There is no longer a cookbook recipe to solving a challenge,” stated Bobby. “You need to take a step back and evaluate the problem. You can’t just go to the workflows that worked in the past.”
Laura reiterated how AI helps her embody an innovative approach: “I am always trying to be more proactive to spot behaviors, trends, and potential risks. AI has been the most useful tool in helping me do that.”
“And highlight your capabilities” to colleagues as well as clients, added Josh. “That’s what will get you the most interesting challenges. If you make connections and go broad, you will have a lot of internal customers.”
The legal industry is just beginning its AI journey. If they’re willing to push boundaries, e-discovery and data professionals will be at the forefront of that transformation.
Graphics for this article were created by Natalie Andrews.