Notwithstanding frequent “hallucinations” and the consequent embarrassment of getting caught relying on court rulings that do not exist, generative artificial intelligence is nevertheless fast becoming a familiar component in the workflows of busy litigators. The reason is that, for many routine litigation chores, generative artificial intelligence technologies are really, really good.
One area where AI shines is the task of writing summaries of deposition transcripts. Transcript review is a critical, time-consuming, and often overwhelming part of busy litigation practices due to the sheer volume and complexity of deposition transcripts. Yes, there is a ten-deposition limit on depositions in federal courts. That limit, however, is frequently exceeded either by stipulation among the parties or by leave of court.
With generative AI’s deposition transcript summarization capabilities, law firms can significantly reduce the manual effort and time typically expended on this routine litigation task.
It’s not uncommon for litigators to take dozens of depositions in major cases, creating thousands of pages of deposition transcripts that must be studied, summarized, categorized, and assigned a role in a party’s litigation strategy. Managing court deadlines, ensuring accuracy, and centralizing the resulting legal work product are persistent pain points for legal teams.
Enter generative artificial intelligence. When assigned the task of summarizing one or any number of deposition transcripts, AI can:
- significantly cut the time needed to write deposition transcript summaries
- create valuable “first drafts” of deposition transcript summaries
- extract key events and write chronologies from data buried across multiple deposition transcripts
- allow litigators to “search” deposition transcripts with concept-based queries rather than traditional keyword-based searches
- accelerate understanding of the case and preparation for dispositive motions and trial
The end result is that, with generative AI’s deposition transcript summarization capabilities, law firms can significantly reduce the manual effort and time typically expended on this routine litigation task. There are other benefits too. When transcript summaries are stored in a centralized location, they can enhance efficiency, consistency, and collaboration among legal teams, allowing litigators to focus more on strategy and less on repetitive tasks.
AI Aids But Doesn’t Replace Lawyers
Despite these advantages, human validation of AI-generated work product remains essential for several reasons:
- Professional Ethics. Litigators have an ethical duty to use AI when it benefits their clients, to be mindful of the risks, to supervise the use of generative AI in their practices, to communicate the use of AI to clients, and to carefully review AI output for misstatements or fact or law. We recently summarized the views of several state bar organizations on the ethical use of generative artificial intelligence in law practice. Another good resource is the American Bar Association’s ABA Formal Opinion 512 (Generative Artificial Intelligence Tools), which discusses in detail how the ABA Model Rules of Professional Conduct apply to generative AI.
- Technology Shortcomings. Generative artificial intelligence tools, while impressive, are still in their early stages of development. They occasionally make mistakes, and they may not have access to all case-related information prior to generating a deposition transcript summary. There is also the danger that AI will miss critical information or misclassify information. Review and reflection by litigation team members are necessary to use these tools effectively.
- Litigator’s Judgment. Generative artificial intelligence technologies cannot (yet) replace the judgment of an experienced litigator. The significance of facts and trends unearthed by AI review may be overlooked by these tools.
While the cost and time needed to learn how to use generative AI are significant barriers to adoption today, many law firms are investing in ongoing training to educate attorneys about AI’s capabilities, limitations, and ethical implications. Continuing legal education programs addressing common AI deployment issues are widely available. Larger law firms with the resources to do so are offering AI “labs” for hands-on experience, creating prompt libraries, and providing tailored training and support to their partners and associates. They believe that early adoption of this technology, even in smaller cases, will help their litigation teams build familiarity with AI and a readiness to deploy it effectively for larger, more complex matters.
Education aside, what’s critically needed now is patience and a persistent desire to drive AI technology adoption forward in a balanced manner that leverages AI for efficiency while maintaining rigorous human oversight. Artificial intelligence technologies promise to transform deposition transcript review, but their greatest value lies in augmenting — not replacing — the expertise and judgment of litigation professionals.
