Skip to content

Our Privacy Statement & Cookie Policy

All Thomson Reuters websites use cookies to improve your online experience. They were placed on your computer when you launched this website. You can change your cookie settings through your browser.

AI

Unlocking the potential of AI for English law

The University of Oxford is undertaking a wide-ranging investigation into the role played by AI in legal services. Oxford’s John Armour and Richard Parnham report from the project’s March kick-off event.

Artificial intelligence (AI) has the potential to transform the legal profession, legal education, and the way that legal services are delivered. Just how extensive these transformations will be was the focus of a two-day conference, held at the University of Oxford’s Saïd Business School on March 18-19.

The conference marked the kick-off of a £1.2 million research project into AI and legal services at the University of Oxford, funded by UK Research and Innovation (UKRI) as part of the Next Generation Services stream of the Industrial Strategy Challenge Fund. The UKRI funding is supporting five interlocking strands of research being undertaken at the university. Collectively, the various components aim to “unlock AI’s potential for English law.” These five strands will focus on a range of issues, including how AI will:

  • impact the business models for delivering legal services;
  • aid dispute resolution, and impact constitutional principles and the design of the justice system;
  • assist with the development of legal research, data extraction/analytics, and reasoning tools;
  • evaluate the possible impact on the skills composition and labor needs in relation to high-value services; and
  • explore the design and delivery of educational programs to address skills gaps that may constrain the deployment of AI solutions within the legal sector.

The goal of the conference was to launch a dialogue with stakeholders about the project’s research questions and build a community with which the research team can continue to engage during the project’s evolution.

Reflecting the research project’s interdisciplinary nature, representatives from across the university outlined their research plans at the launch event, including members of the Law Faculty, Economics, Computer Science, and Education Departments, and the Saïd Business School.

A distinguished panel of external experts also shared their views, including The Rt Hon Lord Briggs of Westbourne, Justice of the Supreme Court; The Rt Hon Lord Keen of Elie QC, Advocate General for Scotland and House of Lords spokesman for the Ministry of Justice; legal technology consultant and author Professor Richard Susskind OBE; and Christina Blacklaws, President of the Law Society of England and Wales. In addition, a wide range of representatives from the worlds of law and computing also offered their insights and experiences. Speakers included Robin Dicker QC of South Square Chambers; representatives of Innovate UK, DeepMind, international law firms Slaughter and May, and Allen & Overy LLP; legal technology leader Thomson Reuters and legaltech startup LexSnap. These were complemented by a wide range of academic perspectives, including from members of the Oxford Internet Institute, the Alan Turing Institute, MIT, University of Bristol, and Duke Law School.

AI still in its infancy in legal

Throughout the proceedings, several speakers made it clear that the use of AI was very much in its infancy in the legal sector, and many of the practical implications of the technology’s development and adoption have yet to be worked through. On the developmental side, it was suggested that the advent of AI may impact on the nature of legal sector recruitment, potentially with an enhanced emphasis on recruiting candidates with backgrounds in science, technology, engineering, and mathematics (STEM) subjects. And, on the deployment side, it was envisioned that lawyers and computer/data scientists would need to work together far more often than previously. This, it was suggested, might not be an easy partnership, in light of law’s inherent ambiguity and the scientific community’s preference for clarity and precision.

At the organizational level, it was proposed that emerging AI tools were likely to impact the way that legal services were delivered and charged for — even the structures of legal practices themselves.

Taking a more philosophical direction, several speakers posed questions about how legal AI might impact the justice system and wider society. For example, how should the justice system be redesigned to take advantage of the potential efficiency improvements offered by AI in a way that nevertheless recognizes the importance of judicial independence?

Would the public accept “disintermediated justice” when issues of personal liberty were at stake — and would such systems be capable of explaining their judicial reasoning? Examples were offered of how the deployment of machine learning-based AI technology in the wider economy was already starting to challenge long-standing legal concepts, requiring new regulatory and economic frameworks be developed in response. One of these examples focused on the dynamic pricing of goods and services, which had the potential to give rise to cartel-like behaviors — but driven by machines, not humans.

At the end of the two-day event, attendees were left with no doubt about the scale of the challenges and possibilities that AI posed to the legal sector, at both a conceptual and practical level. At the same time, there was also a real sense that — collectively — the University’s research project would play an important part in responding to those challenges and opportunities. This would be made possible by various research disciplines working together — often for the first time — along with an extensive engagement with industry, regulators, and other key stakeholders.


Professor John Armour is Professor of Law and Finance at the University of Oxford; and Richard Parnham is a Postdoctoral Research Fellow at the University of Oxford’s Saïd Business School.

More answers