Generative artificial intelligence poses significant implications for legal work and tasks, recent studies suggest. A study conducted by a team of professors from the University of Southern California and University of Minnesota showed substantial efficiency increase in the completion of routine legal assignments with the aid of generative AI. This increment in work speed reportedly amounts up to 32%.
However, the study’s findings were not entirely unilateral with respect to work quality. It found that the effectiveness of generative AI in improving the quality of completed tasks frequently varied and pivoted on factors like the individual skill level of legal professionals and the nature of the law-related tasks at hand.
The researchers behind the study referred to it as the first randomized controlled trial concentrating on the effects of large language models on legal analysis.
To better understand the results, it’s of note that the study focused on the role of GPT-4, a highly advanced AI entity, when assisting with four tasks commonly encountered by junior lawyers in the initial years of their career.