In a move reflective of the evolving intersection of law and technology, eDiscovery Assistant, a reputed legal research database, unveiled its pioneering generative artificial intelligence (AI) integration this past Monday.
The woman behind the innovation, Kelly Twigger, the helm at ESI Attorneys in Colorado and eDiscovery Assistant’s CEO, imparted to Legaltech News that this innovative step primarily aims to provide concise case summaries. Yet, she hinted at a broader horizon, mentioning, “the coming months will see more use cases.”
The Breakdown: Empowering legal professionals with efficiency, the generative AI offers precise summaries derived directly from eDiscovery Assistant’s extensive database. Twigger elaborated that these summaries would encompass a “one to two sentence description” focusing on ESI (Electronically Stored Information) specifics in any given case. She also highlighted the platform’s feature allowing users to enhance search accuracy with “issue tags,” such as “spoliation,” “sanctions,” and “proportionality.”
Marking their initial exploration of generative AI, eDiscovery Assistant has plans for more applications beyond mere summarization.
The Rationale: Twigger outlined twin catalysts for this integration: the ever-growing demand from users and the skyrocketing costs associated with legal research. She sees this AI-driven tool as a streamlined means for attorneys to remain abreast with pervasive ESI challenges. Twigger further emphasized the economical aspect, stating that the spiraling research costs have curtailed clients’ willingness to pay repetitively for similar insights. Consequently, these AI-summaries present an efficient alternative, offering a snapshot without necessitating a full case review.
Technical Specs: Leveraging OpenAI’s renowned GPT-3.5 model, eDiscovery Assistant has solidified its summary capability. Twigger, however, anticipates further enhancements with the introduction of newer GPT iterations, particularly eyeing GPT-4. She assuaged potential data security concerns, clarifying that summaries are generated solely from specific decisions. “The only contents that we’re pulling from for this summary is the publicly available information within the decision. There’s no commentary,” she asserted.
The Competitive Landscape: The past year has witnessed a surge in legal platforms, with giants like LexisNexis and Thomson Reuters, adopting generative AI. Yet, for eDiscovery Assistant, the distinguishing factor remains their unparalleled specificity tailored to e-discovery nuances. Twigger articulated this by elucidating the depth of their summaries, which offers insights far beyond generic legal outcomes. Achieving such granularity, as Twigger shared, required exhaustive “10 to 12 rounds” of prompt engineering per case.
As AI continues to permeate the legal sector, eDiscovery Assistant’s strategic move underscores the industry’s commitment to harnessing technology for precision and efficiency.
