As lawyers, we often bristle at the idea of breaking our work down into a series of tasks that can be delegated or even automated. We prefer to think of our jobs as requiring a high degree of intellectual skill, judgment, and artistry — something that cannot be replicated by artificial intelligence. However, the recent launch of AI-driven features by Thomson Reuters, including those geared toward legal research and document analysis, invites us to confront the complex, often unacknowledged, inner workings of a lawyer’s mind, and to consider how technology might be able to augment or emulate these processes.
As reported by Above the Law, these new additions function as “skills” — a term meant to draw attention to the intricate tasks involved, such as building deposition outlines based on the nuances of the case and underlying legal standards or flagging discovery documents that may potentially indicate fraud.
What becomes clear in working with AI is that it invites us to think critically about our professional practices. With technological enhancements, each step in the legal process can be broken down into various tasks, each requiring its unique problem-solving approach. Understanding this can not only help improve the functionalities of AI such as Thomson Reuters’ but also help us become more aware of our work methodologies.
In essence, the challenges of integrating AI into legal practice go beyond the risk of AI ‘hallucinations’ where the system might fabricate false cases. Experts like Mike Dahn, Head of Westlaw Product Management, believe retrieval augmented generation, a method that allows large language models to draw on trusted data, mitigates these risks. The hard part, however, is achieving accuracy within reality’s bounds, something young lawyers grappling with case complexities would empathize with.
However, that’s not to say AI cannot add value. The AI research window allows users to pose a simple question and receive a meticulously researched answer, a process that could traditionally take a junior associate hours to compile into a memo. But the question remains: What if the ‘right answer’ isn’t what the attorney needs?
In many cases, especially litigation, research aids in finding not just the ‘right answer’ but also exploratory paths that can give their clients an edge. Currently, AI technology may not be best suited for this ‘higher level thought process.’ But as we continue to understand our own processes better, it is likely, given the rapid advancements in tech, that this reality might not be too far.
As AI makes strides in the legal industry, it inevitably forces us to reckon with our mental processes, pushing us to understand ‘how we do what we do.’ This journey of self-discovery could well be the beginning of a new phase in legal practice, one where AI doesn’t usurp lawyer jobs but rather, enhances them.