The general consensus for legal AI companies and practicing lawyers like Professor Eric Talley of Columbia Law School is that AI could reduce contract review time from 40 minutes to 2 minutes, according to Keith Porcaro of Wired. Lawyers are swamped (according to the NYT article “A.I. Is Coming for Lawyers, Again”) so having an AI comb through that contract, redline improper language, and identify unapproved clauses could be the future we’re looking at. “
The reason the public has been seeing so much AI content just now is because last November, OpenAI put out a public interface of AI called ChatGPT, which is also technology involved in AI legal companies. Missouri Lawyers Media wrote about what these essentially are: “A large language model is a computer program that has been trained on a vast amount of text data to understand the rules and patterns of language.” Like a person, the GPT‑4 is taking the Bar Exam and passing in the 90th percentile, something that has been a big deal in recent news for these programs. This improvement is unnerving with the knowledge that it will only improve and at a rate much faster than humans.
However, what all these sources point out is that technology is amazing, but not amazing enough yet to know what numbers to calculate and what is important to being a good attorney. As Professor Lawrence Solum of University of Virginia warns us, AI must prove to be reliable for this grunt work as AI has shown before to be faulty, like when ChatGPT showed Talley a case that didn’t exist when asked to find a case with a special claim. Jonathan Sternberg, a lawyer in Kansas City, couldn’t get AI to find examples of Missouri cases he knew existed. As of now, AI is a tool that’s building steam and on the precipice of becoming immersed in our legal system.