The Legal Singularity: U of T Law profs on how AI will make the law 'radically' better
Benjamin Alarie, a professor in the University of Toronto’s Faculty of Law, has long believed artificial intelligence will bring seismic change to the legal profession and, consequently, society – resulting in what he’s dubbed ‘the legal singularity.’
In a forthcoming book, Alarie tackles the topic with Abdi Aidid, who recently joined the faculty as an assistant professor.
The pair argue that the proliferation of AI-enabled technology – and specifically the advent of legal prediction – will radically change the law profession and facilitate “a functional ‘completeness’ of law, where the law is at once extraordinarily more complex in its specification than it is today, and yet operationally vastly more knowable, fairer, and clearer for its subjects.”
Alarie says that’s in stark contrast to how law is practised now.
“There is a ton of uncertainty in the law – we often just don't know what the right legal answer is,” says Alarie, who is the Osler chair of business law. “Uncertainty about facts and law drives litigation. Even if there aren't disputes about the events involved, litigation arises due to a dispute about how the law applies to those facts.”
Alarie and Aidid suggest the book,The Legal Singularity: How Artificial Intelligence Can Make Law Radically Better (University of Toronto Press, 2023), should be of interest not only to lawyers and technologists, but anyone interested in the future of the labour force or social institutions beyond the law.
“The legal singularity reflects the full development of our legal system, becoming more complete and accessible through advanced technology,” says Aidid. “The idea is that once we are able to reduce uncertainty, individuals and institutions will have a real-time sense of their legal rights and obligations.”
Alarie’s interest in legal technology began more than a decade ago, when he served as associate dean of the faculty’s JD program and was tasked with revisiting how the law faculty delivered its first-year curriculum.
“I remember sitting at my desk and thinking, it’s been almost 40 years since we’ve had a major reform to the curriculum,” he says. “What could change over the course of the next several decades?
“I thought about the deep learning work that was being done by [University Professor Emeritus] Geoffrey Hinton in the computer science department [in the Faculty of Arts & Science] here at the U of T, and how computing power keeps doubling every couple of years and is becoming massively less expensive over time.”
He also recalled Oliver Wendell Holmes Jr., a renowned legal scholar and former U.S. Supreme Court justice, once offered a provocative view that law is all about prediction. Alarie thought to himself: “Well, what is machine learning? It’s a prediction technology. All these ideas were swimming around in my mind: machine learning is about prediction. Machine learning is getting way better. Law is ultimately about prediction.
“I'd better be thinking about how machine learning is going to influence the practice of law, because that's going to have big implications for how we want to teach our students.”
In 2015, Alarie co-founded legal tech startup Blue J Legal with U of T Law faculty members Anthony Niblett and Albert Yoon. The company’s software draws upon AI to provide instant and comprehensive answers in complex areas of tax, labour, and employment law.
Aidid joined Blue J Legal in 2018.
“While I was an adjunct professor here, teaching courses in legal research and writing, I was seeing first-hand the difficulty [with the way] we currently do legal research and was just hoping for a technological solution,” says Aidid, who served as the startup’s director and vice-president of research, and remains with the company as an innovation specialist.
“Being able to help build Blue J and contribute to improving a profession that I care deeply about was really appealing to me.”
The authors argue that the legal profession has so far failed to keep pace with other industries and professions.
“If you were transported back 50 years into a law school classroom, or a courtroom, it would look largely the same as it does today,” says Aidid. “There might be a laptop on someone's desk, but by and large, we're doing the same things. It’s not just about tech adoption – it's about changing some of the core assumptions about what it means to be a good lawyer, legal academic and a good law student.”
That includes how lawyers bill their time.
“If you come to me and you ask me a legal question, I might have an instinct about the answer but in order to give you professionally sound advice, I'm going to go off and do research until I feel fairly certain about my advice,” says Aidid. “But with the advent of machine learning, you're able to quickly synthesize all the case law in a matter of seconds.”
Alarie says future technological innovation that can interpret legislation, legal principles and translate them into appropriate legal guidance will result in better legal decisions for society. He says it’s not meant to supplant or replace legal professionals, but to enable them to provide fairer and more informed decisions. “For example, it would provide judges with more information to better exercise their discretion,” Alarie says.
He adds that that the notion of a legal singularity is best regarded as an ongoing process of improvement, rather than a final destination. Aidid, for his part, emphasizes the important role lawyers will play in making sure legal technologies are “designed appropriately, ethically and effectively.”
One thing is clear, the authors say: the application of AI to the law is no longer a fanciful sci-fi thought experiment.
“How we get from here to there – wherever that there is – could be a bumpy ride,” says Alarie. “Our goal is to really spread these questions – about the legal singularity – as widely as possible because we don't think they have easy answers, but they are important questions.”