Will Artificial Intelligence Have a Radical Impact on the Legal Profession?

Artificial intelligence (AI) has been a hot phrase for lawyers, both in-house and in private practice, for some years. But will it fundamentally transform the day-to-day work that lawyers do? As part of Thomson Reuters ongoing Legal Debate Series in the UK, domain experts from around the Legal industry set out to test the hype with the motion: “Artificial Intelligence (AI) will fail to have a radical impact on the legal profession.”

In front of a packed house in Thomson Reuters Canary Wharf auditorium in London, Reuters Editor-At-Large Axel Threlfall chaired of the debate and set the scene by offering some potential radical impacts of AI, including the elimination of thousands of junior and paralegal roles to be replaced by AI or the use of AI to offer client-facing front-end services. Might chatbots soon provide low-cost legal advice? Could the advent of emotionally intelligent AI render the very role of a solicitor or barrister redundant?

The first round of audience polling revealed that just 32 percent agreed that AI would fail to have a radical impact on the legal profession. Forty-four percent disagreed, leaving 24 percent undecided for our debaters to win over.

Mark Edwards, vice president at Rocket Lawyer, took to the podium first, arguing for the motion. His company provides online legal services to small business and consumers. He was clear that “legal services are transforming,” and will be “significantly different in ten years’ time.” However, he also argued that AI itself wouldn’t radically impact the profession anytime soon – certainly not within the next 25 or 30 years. He cited a recent study that suggested less than 30 percent of a lawyer’s daily tasks could be automated. Although Moore’s law dictated that machines would double in processing power every 18 months, this only guaranteed that they’d be faster, not smarter. Technology and other drivers would help the legal industry to transform significantly, but the impact of AI on this wouldn’t be radical.

Dr. Peter Waggett, formerly a rocket scientist and now Emerging Technology Programme Leader at IBM Watson, argued against the motion. He urged us to consider the positive impact the development of AI could have on skills shortages in areas like medicine and the law. AI systems “never get tired, never sleep, and they continue to improve.” He drew parallels between the signals of impending transformation in the rocket science industry and the current climate of the legal industry. He also pointed out the thorny issue of liability: when AI takes control of systems, like driverless cars, who is liable when things go wrong?

Andrew Bodnar, a barrister at Matrix Chambers, supported the motion and noted two important questions: “Is it realistic to expect at any time that a computer could do what we as lawyers do? And even if it is, do we want them to?” He noted that AI is already helping to transform the process work that lawyers undertake, but it would struggle to achieve the understanding that a human lawyer can bring to the table – knowing not just what the law is, but why it is the way it is. Moreover, would ethics prove too big a hurdle for AI to surmount? He concluded that on its current trajectory, AI may well make lawyers more efficient, and make their services more accessible, but it won’t radically alter what they do as lawyers.

Edward Chan, a partner at Linklaters, argued against the motion. He wondered what sort of impact would be considered radical for the profession: barristers in court may not be greatly affected, but solicitors reviewing documents could see a substantial impact. He noted that the legal profession, and the pyramid-shaped structure of law firms, are still to a large extent shaped by the advent of word processing. The automation of basic tasks could change the shape of that pyramid fundamentally and challenge its pricing structure. Clients are used to paying on a time basis, and machines may handle tasks in a radically different time frame.

The audience was invited to put their questions to the panel and raised many insightful points, such as the relationship between AI and big data; and the question of how high up the value chain AI would be allowed to operate in a law firm. One audience member asked what young lawyers should start thinking about as they join an AI-influenced workforce – do they all need to learn to code? As the Big Law representative, Chan emphasised that legal reasoning skills are still prized above all others, as ultimately the machine will need to be taught by the lawyers.

The audience voted for the final time, and a significant number of undecided voters had moved to agree with the motion that AI would fail to have a radical impact on the legal profession. The final tally showed 53 percent in favor and 46 percent against the motion, with just one percent undecided, meaning the motion was carried.

Please join us for our next legal debate on Sept. 1, to discuss accountability in law for promises made during election campaigns. Who should be responsible amongst campaigners, parliament and the media? And who should hold them to account? Follow @TRLegalUKI for updates, and visit our Innovation in Law microsite here.

You can also watch the debate in full below.

This post was written by Tom Bangay, content manager with Thomson Reuters Thought Leadership team in the UK.

Alex Cook

Leave a Reply