Robert Keeling Quoted in "What the Legal Battle Between the New York Times and OpenAI Means for E-Discovery"

In discussing how the case raises the question of how equipped parties are to handle the preservation, production, and use of GenAI data, the article mentioned the recent Georgetown Law Advanced eDiscovery Institute, where members of a panel focusing generally on the intersection of GenAI and eDiscovery noted that preserving, producing and parsing GenAI data such as inputs, outputs and system data present novel challenges as the field grapples with how to manage new data types.

Robert Keeling, a partner at Redgrave, noted that, unlike emails, which tend to stay grouped in chains, prompts and outputs often show up as separate, disaggregated documents in discovery.  This issue is compounded by the way gen AI models produce “artifacts,” byproducts through which they essentially show their work.  Together, these factors mean that gen AI systems produce unprecedented amounts of data, much of which can be confusing and hard to place in context.