Continuing Conversations from the Colloquium (Part 1): Robots vs Lawyers

In the “Continuing Conversations from the Colloquium” series, we look at some of the key themes arising from the Law Society’s Colloquium on ‘The Role of Lawyers in the Age of Disruption: Emerging Regulatory Challenges’, which was held as a live webinar on 19 May 2020. One recurrent theme, as seen from participants’ questions submitted during the discussions for Panels 1 and 2, was the popular Robots vs Lawyers debate. In the first of a two-part series on “AI & Lawyers”, we explore some interesting issues on whether robots will take over lawyers based on a curation of our participants’ questions.

This brief note is written by Alvin Chen, Director of the Legal Research & Development department at the Law Society of Singapore.

First of all, let me thank all the 326 participants of the Colloquium for taking the time to spend most of their working day on 19 May with us to explore important issues relating to the future of lawyers. Due to the tight schedule of the webinar, the panels were unable to address many of the excellent questions posed by the audience. But, we hope to address some of the common themes arising from these questions in this “Continuing Conversations from the Colloquium” series. The thematic issues regarding the Robots vs Lawyers debate that are explored below are based on an edited version of participants’ questions. You are welcome to contribute further thoughts on these issues by writing to the Legal Research and Development department at lrd@lawsoc.org.sg.

Robots taking over lawyers seems to be far-fetched. Isn’t artificial intelligence (AI) merely a tool for lawyers to use just like legal precedents, statutes or law textbooks?

My take: The difference between AI and other legal resources can be explained through the concept of “task encroachment” that Daniel Susskind refers to in his recent book A World Without Work: Technology, Automation and How We Should Respond. The central point in the Robots vs Lawyers debate is not whether we can create robots that can talk like lawyers, but whether certain tasks that lawyers now perform can be outsourced to AI. Daniel Susskind argues that our manual, cognitive and affective capabilities may be taken over by robots in the future. For example, lawyers’ cognitive capabilities (e.g. in advising the client on the likely outcome of his or her court case) may, to some extent, be replaced by the predictive powers of AI software in the future. This is something that pure legal resources like legal precedents, statutes or law textbooks cannot do on their own.

If AI can take over some of the tasks currently performed by lawyers (especially junior lawyers) to an acceptable level of accuracy, how will this impact on their training in future?

My take: A recent Law.com article suggests that the possibility of “skills erosion” of junior lawyers is a real concern. For law practices are that already using AI software to perform some of the tasks traditionally done by junior lawyers, there seems to be an inevitable trade-off between achieving the speed and efficiency that clients may demand, and giving enough opportunities for junior lawyers to learn and develop their professional judgment. As the article observes, there is also a risk management aspect in that inadequate skills acquisition in the age of AI may have catastrophic consequences in the future if junior lawyers become more susceptible to errors.

If this problem becomes widespread, legal industry stakeholders, and not only law practices, should consider whether there are other avenues for junior lawyers to gain the necessary skill-sets if they are unable to do so within their law practices.

Will the use of AI in the legal profession come to a point when lawyers will be reduced to providing inputs for the superior AI software only?

My take: From a broader perspective, this is a concern that has been characterised as a “doomsday scenario”. No one can say for sure if we are on an inexorable path towards a machine take-over, although some leaders and pioneers in the AI industry have sounded alarm bells that we are already on the road to the destruction of humanity. On the other hand, some commentators foresee AI and humans co-existing in a collaborative way. For example, a recent article in The Straits Times suggests that AI will not displace humans by a long way, although some jobs will necessarily be lost. This is simply because the human factor is too important to be displaced.

Lawyers are just a sub-set of this larger debate, but there are good reasons to be concerned given that the legal profession, in view of its lack of technical expertise, is not in the driver’s seat in driving AI and it is unclear what outcomes we are seeking for the legal profession as a whole.

The LawGeex AI study, where AI outperformed lawyers in accurately reviewing non-disclosure agreements, has often been cited as an example of AI having the potential to surpass lawyers. But would a disproportionate amount of resources be required to generate an AI software to review more complex contracts?

My take: This is really a question for the techies, but the thrust of the question goes back to what I mentioned previously, i.e. what outcomes are we seeking for the legal profession? The assumption seems to be that if we can perform legal tasks faster and more efficiently by using AI, it will be more optimal for lawyers and clients. To some extent, that assumption may hold true for mundane legal tasks. But as we ascend the ladder, more difficult questions may need to be answered. Will more jobs be put at risk? Who should decide on the objectives that AI should be used for? How much resources will be invested in developing AI, and if such resources are disproportionate to AI’s probable success rate (or benefit), why should we invest in AI in the first place?

As lawyers, we may not have the answers to all these questions, but it is timely to think more carefully about the wider implications of AI, not only on the legal profession but also beyond it.

Watch this space for Part 2 of ‘AI and Lawyers’, where we will examine more interesting questions that were raised at the Colloquium.