Every week, the Array team reviews the latest news and analysis about the evolving field of eDiscovery to bring you the topics and trends you need to know. This week’s blog covers the week of February 5-11. Here’s what’s happening.
Cases involving emojis continue to rise: eDiscovery implications
As communication evolves, so do our needs to correctly capture and catalog it for litigation. The latest frontier? Emojis! (Insert smiley face here)
Indeed, emojis – cartoon images that can be inserted into texts, emails and messaging – have made their way into recent casework as relevant context. Eric Goldman, Associate Dean for Research and a Professor of Law at the Santa Clara University School of Law, tracks cases involving emojis on his blog and found 216 emoji-related cases in 2023, up from 177, 163, 125 and 98 over the last four years.
One of last year’s cases involving emojis, covered by Doug Austin at eDiscovery Today: A U.S. District Court judge found a social media message sent by a Bed Bath & Beyond investor containing a moon emoji was actionable. Said the judge: “Some online communities understand the smiley moon emoji to mean ‘to the moon’ or ‘take it to the moon.’…In other words, according to Plaintiff, (the investor) was telling his hundreds of thousands of followers that Bed Bath’s stock was going up and that they should buy or hold. They did so, sending the price soaring.”
Beyond the contextual quandaries, emojis bring practical challenges for eDiscovery. As Kathryn Cole at Greenberg Traurig’s eDiscovery Watch notes:
- Different systems support different emojis; an image sent from an iPhone may render differently (or not at all) for an Android.
- Some instant messaging systems have proprietary emojis and allow users to create their own.
- Emojis often evolve; for example, one operating system changed a pistol emoji to a less-dangerous water gun. (But, when seen on a different platform, the image may still render as the original.)
- The meaning of symbols can vary by culture; for example, the “thumbs-up” emoji can be considered vulgar in some parts of the world, yet is often a straightforward authorization in the U.S.
In the past, emojis may have been considered innocuous. With today’s communication style that is often dominated by acronyms, GIFs and memes, lawyers will need to pay closer attention to what is being said or implied when emojis are used in conversations to understand the potential true context of the messages.
Search terms vs. LLM prompts
Joe Patrice at Above the Law writes about the parallels and differences between Boolean search construction in the early days of eDiscovery and today’s prompts for Large Language Models.
While constructing the right search terminology in the past could lead to predictable results, today’s prompt engineers need to consider what part of their prompts AI is paying attention to and how that affected the computer’s output.
“Boolean had the advantage of being, more or less, a fixed language across tools,” Patrice writes. “Large Language Models keep all that stuff behind the curtain of its natural language, chatbot-inspired interfaces. Depending on the precise situation, the model might be training itself how to react to prompts and coming up with idiosyncrasies that no one expected.”
Responsible use of AI dictates that all AI-driven results should be verified; certainly at these early stages when we are still trying to fully understand what is “behind that curtain.”
Preparing for data deletion
Matt Kelly at Radical Compliance writes about the mechanics of complying with a FTC data deletion order faced by a South Carolina technology company: Within 90 days, the company must destroy all customer data it isn’t using and must not keep customer data beyond the point it needs it.
Some considerations, Kelly writes:
- Inventory your data: What files contain which types of data (names, phone numbers, birthdays, etc.) and where are these files located?
- Cross-reference data: Even if the company isn’t using some of the customer data, could some of it be subject to a litigation hold?
In these situations, a Data Map is key. Organizations should have one on file that is kept updated as features change or new technologies are adopted and old solutions are deprecated. Understanding where data is stored, the types of information it could contain and the business and IT leaders responsible for the maintenance of these systems could greatly simplify such onerous requests.
Other recent eDiscovery news and headlines:
Native Form Production Denied After PDF Files Produced (eDiscovery Today)
