The legal profession has been quick to embrace AI. Law firms and corporate legal departments are using AI for repetitive tasks such as contract and document review. Courts are exploring AI for workforce management, processing information and creating analytics to predict the outcome of cases.
Just as some lawyers were wary of the transition from paper ledgers to online databases, many worry about how GenAI will change the traditional practice of law. Professor Cristina Knolton, who teaches legal analysis and writing at Southwestern Law School in Los Angeles, is leading several initiatives to incorporate AI into the school’s curriculum. Knolton expects lawyers will be able to streamline some tasks, which could reduce billable hours.
“If there is a faster, quicker way to do your research, and it’s reliable and accurate, and you understand how to check that accuracy, we should encourage students to learn that, because that’s what the real practice is going to be like,” she said. “They’re going to have to limit their bills for their clients. You’re not going to be able to charge to do your traditional research if GenAI becomes traditional.”
Legal database stalwarts LexisNexis and Westlaw are creating proprietary AI designed to draft, summarize, research and cite legal documents with large language models, which are anticipated to be more secure than ChatGPT.
These improvements are essential if lawyers are to maintain client confidentiality and data privacy.
“Right now, if you put information into ChatGPT, it trains the model. It’s out there in the universe,” explained Knolton. “So lawyers aren’t able to just search for something or have a draft, a memo, or even an opening or closing statement, because to give it any real facts and real specifics about your client’s case is a violation of your duty of confidentiality.”
A deluge of AI tools for law firms — including automation technology from startups such as LegalMation, EvenUp and PaxtonAI — are flooding the marketplace. Some major law firms are developing their own in-house GenAI.
Stetson University’s Davis recently used a springtime sabbatical to study how GenAI will affect legal communication. In addition to her work as a law professor, she’s the founding director of Stetson’s Institute for the Advancement of Legal Communication.
“Generative AI is really interesting from the perspective of creating something from scratch,” she said. “It is doing that writing activity that we typically assign to a human actor and with the apparent ease of producing that content as a human actor does.”
Law schools would be remiss to ignore the implications of GenAI on legal education and the profession.
“I can’t think of something more revolutionary to happen to writing … since probably the arrival of the personal computer, right when we stopped using typewriters,” Davis said. “It’s the newest thing I’ve seen in my career that has the potential to fundamentally change the way we think about what it means to write in the law.”
GenAI systems such as ChatGPT can be a liability in the hands of an unskilled lawyer, Davis said. She points to the case of a New York City law firm that used ChatGPT to create a court brief that cited fake federal cases. Lawyers for Levidow, Levidow & Oberman incorrectly assumed the chatbot was a sophisticated search engine. The misstep cost the firm a $5,000 court-ordered fine and professional embarrassment.
“The large language models are engaged in prediction,” Davis explained. “They’re not ‘thinking’ about things in the way you and I are thinking. They are thinking about what language makes sense. They’re thinking about what words go together based on the patterns of words they’ve identified in the vast amount of language [they’ve] analyzed.”
To be sure, GenAI is not a fact-checker.
“It is identifying patterns and language. It’s not verifying,” Davis said, explaining that the software runs statistical computations based on vast amounts of input data. “I say, it turns words into math and math back into words.”
The tendency of current GenAI to “hallucinate” or seemingly pluck some data out of thin air means lawyers still need to have their heads in the game.
“The obligation of the ethical lawyer who uses any of these tools is going to be to ensure that whatever they endorse as their own writing, if it’s generated by a large language model, that they’ve checked the analysis,” Davis said. “They need to be confident that what they’re providing is as accurate as if they had written it themselves.”
For all the technological advancements coming into use, the fundamental skills of lawyering remain unchanged, Davis said. Attorneys still need to ask relevant questions, identify important issues, devise good arguments and counterarguments, assess client needs, and understand the principles of negotiation and mediation.
“What might change is the way in which you get to produce the documents of practice,” she said. “Over time, we may see that more and more of the production part of that can be done with the help of a really sophisticated tool, but none of that will ever replace the foundational skills that they’re learning in law school.”
Her message to law school students navigating new technologies: “You need lawyer intelligence to use artificial intelligence.”
How AI can help lawyers
Thomson Reuters recently surveyed legal professionals, asking in what areas AI had the most potential to help and/ or hinder law firms.
The results, published in its August 2023 Future of Professionals Report, showed that the majority expected AI to have positive effects in most areas, perceiving it as a catalyst for growth.
“Legal professionals in both law firms and corporations see AI as having the ability to enable growth opportunities,” the report said. “For law firms, AI reduces friction to launch new services or expand into new markets.
“For example, AI potentially can free up time for professionals to identify and evaluate attractive markets for expansion. For corporate departments, AI can free up their time to dedicate to supporting their organization’s strategy and growth objectives.”
The report recommended the following:
Embrace lifelong learning and upskilling, and seek to develop a combination of technical, cognitive and social skills that will enable you to work effectively with AI.
Define the role and scope of AI in your work. Identify the tasks that could be automated, augmented or delegated to machines and the tasks that require human oversight, judgment and creativity.
Consider the ethical, legal and social implications of AI. Adhere to the principles of responsible and ethical AI use, such as fairness, accountability, transparency and privacy.
Read (How AI is changing lawyering) for more content.
