Examining the Intersection of Artificial Intelligence and Law in Providentia’s Latest Series
Article
February 6, 2024
In “AI and the Law,” the second installment of Providentia’s Future of Law Series, legal pioneer Richard Susskind, mediator David Ryan of Providentia and Mark Ross from Deloitte Tax LLP explored the evolving narrative of artificial intelligence (AI) within the legal realm, exploring the history of AI and its increasingly significant role in shaping the future of legal services.
Susskind, whose passion for AI was ignited during his time at Glasgow University in the ’80s, traced the trajectory of legal technology from its initial 65 years of automating processes to the recent shift toward innovation with generative AI. “The long term, 2030 and beyond, is going to be the era of innovation,” he said.
He discussed the history of AI from its inception in the ’50s, through its theoretical applications in the ’70s, to its first practical applications in the ’80s, exemplified by the first commercial AI system for lawyers he co-developed. This system, designed to interpret complex legal documents, showcased AI’s potential to surpass traditional legal methods. “For those lawyers who somehow think that AI was invented about 14 months ago… it’s older than me. It’s in its mid-60s now,” Susskind said.
Reflecting on AI’s unexpected advancements, Susskind pointed to the internet and developments in computing power, citing IBM’s Deep Blue, the chess-playing system that defeated Garry Kasparov in 1997. Susskind also discussed the shift from programmed to learning systems, emphasizing two types of machine learning in law: predictive and generative. He illustrated this with examples like Lex Machina, a legal analytics company, and generative tools like ChatGPT, which can be used to draft and summarize legal content.
Expanding on these systems, Susskind explained that tools like Lex Machina can effectively predict legal outcomes despite having no inherent understanding of the law. They leverage vast amounts of objective data from past cases, such as details about judges, parties and the nature of disputes. “It turns out, if you have enough objective data about past cases, then using machine learning technology, really using computation statistics, you can make predictions in some cases that actually outperform the legal method,” he said. This capability raises questions about the traditional legal approach, particularly when clients primarily seek outcomes, not methodologies.
Furthermore, Susskind touched on the application of AI in tasks like due diligence and document review. Machine learning systems, once trained, can predict which documents a corporate lawyer might deem significant in a deal or isolate important documents in litigation. This shift from theoretical applications in earlier decades to practical, machine learning-based systems at the turn of the century marks a significant evolution in legal technology.
“I think ChatGPT was the single most remarkable development I’ve seen in more than four decades,” Susskind said, cautioning against underestimating AI’s future potential based on its present capabilities. He stressed that generative AI marks a new chapter in AI’s history, which is still in its early stages.
Highlighting AI’s rapid evolution, Susskind predicted that advancements like ChatGPT could soon impact some roles, such as his as a legal columnist. “I can see a way in which a big chunk of my work will no longer be available to me, not in 20 years’ time, but within five years’ time,” he said.
Susskind prophesied a future where AI-empowered in-house legal departments could challenge traditional law firms, democratizing legal work. He also predicted increased importance for roles such as legal no-coders, engineers, technologists and data scientists, arguing that these professionals are the lawyers of the future.
“This always gets lawyers upset because these people didn’t go to law school; they don’t understand the traditional advice. And I say that’s missing the point,” Susskind said. “What really matters to clients is having their problem solved. And if their problems are solved in the future by systems designed by these talented professionals, then I’m happy to call them lawyers, if only for a discussion.”
In his conclusion, Susskind likened the necessity for lawyers to adopt technology to surgeons staying current with medical advances. He asserted that embracing technology is crucial for representing clients’ interests effectively and expanding access to legal services.
Mark Ross from Deloitte added a practical dimension, discussing Deloitte’s research on generative AI in enterprises. Despite the excitement around AI, Ross noted its slower adoption in legal functions compared to other sectors, highlighting an opportunity for disruption, especially in contract management.
The program concluded with Susskind advocating for a proactive, adaptive approach among legal professionals. He encouraged embracing a research and development mindset and experimenting with new technologies, reflecting his vision for a legal industry that is innovative, adaptable and prepared for the challenges and opportunities presented by generative AI.
Susskind, whose passion for AI was ignited during his time at Glasgow University in the ’80s, traced the trajectory of legal technology from its initial 65 years of automating processes to the recent shift toward innovation with generative AI. “The long term, 2030 and beyond, is going to be the era of innovation,” he said.
He discussed the history of AI from its inception in the ’50s, through its theoretical applications in the ’70s, to its first practical applications in the ’80s, exemplified by the first commercial AI system for lawyers he co-developed. This system, designed to interpret complex legal documents, showcased AI’s potential to surpass traditional legal methods. “For those lawyers who somehow think that AI was invented about 14 months ago… it’s older than me. It’s in its mid-60s now,” Susskind said.
Reflecting on AI’s unexpected advancements, Susskind pointed to the internet and developments in computing power, citing IBM’s Deep Blue, the chess-playing system that defeated Garry Kasparov in 1997. Susskind also discussed the shift from programmed to learning systems, emphasizing two types of machine learning in law: predictive and generative. He illustrated this with examples like Lex Machina, a legal analytics company, and generative tools like ChatGPT, which can be used to draft and summarize legal content.
Expanding on these systems, Susskind explained that tools like Lex Machina can effectively predict legal outcomes despite having no inherent understanding of the law. They leverage vast amounts of objective data from past cases, such as details about judges, parties and the nature of disputes. “It turns out, if you have enough objective data about past cases, then using machine learning technology, really using computation statistics, you can make predictions in some cases that actually outperform the legal method,” he said. This capability raises questions about the traditional legal approach, particularly when clients primarily seek outcomes, not methodologies.
Furthermore, Susskind touched on the application of AI in tasks like due diligence and document review. Machine learning systems, once trained, can predict which documents a corporate lawyer might deem significant in a deal or isolate important documents in litigation. This shift from theoretical applications in earlier decades to practical, machine learning-based systems at the turn of the century marks a significant evolution in legal technology.
“I think ChatGPT was the single most remarkable development I’ve seen in more than four decades,” Susskind said, cautioning against underestimating AI’s future potential based on its present capabilities. He stressed that generative AI marks a new chapter in AI’s history, which is still in its early stages.
Highlighting AI’s rapid evolution, Susskind predicted that advancements like ChatGPT could soon impact some roles, such as his as a legal columnist. “I can see a way in which a big chunk of my work will no longer be available to me, not in 20 years’ time, but within five years’ time,” he said.
Susskind prophesied a future where AI-empowered in-house legal departments could challenge traditional law firms, democratizing legal work. He also predicted increased importance for roles such as legal no-coders, engineers, technologists and data scientists, arguing that these professionals are the lawyers of the future.
“This always gets lawyers upset because these people didn’t go to law school; they don’t understand the traditional advice. And I say that’s missing the point,” Susskind said. “What really matters to clients is having their problem solved. And if their problems are solved in the future by systems designed by these talented professionals, then I’m happy to call them lawyers, if only for a discussion.”
In his conclusion, Susskind likened the necessity for lawyers to adopt technology to surgeons staying current with medical advances. He asserted that embracing technology is crucial for representing clients’ interests effectively and expanding access to legal services.
Mark Ross from Deloitte added a practical dimension, discussing Deloitte’s research on generative AI in enterprises. Despite the excitement around AI, Ross noted its slower adoption in legal functions compared to other sectors, highlighting an opportunity for disruption, especially in contract management.
The program concluded with Susskind advocating for a proactive, adaptive approach among legal professionals. He encouraged embracing a research and development mindset and experimenting with new technologies, reflecting his vision for a legal industry that is innovative, adaptable and prepared for the challenges and opportunities presented by generative AI.
For more information on upcoming sessions of Providentia’s Future of Law Series, please visit the Providentia events page.