含羞草社区

The past few decades have seen lawyers address and effect changes to practice methodologies as a consequence of changed technology. That has accelerated since 2018 and the pace of change seems likely to increase. For the purposes of this article, the author uses “artificial intelligence – AI” as commonly applied and ignores the technical aspects of whether or not the utilities available to lawyers as of date are strictly “intelligent”.

The changes to practice that are now necessary differ from those which were effected pre-AI. Pre-AI changes were largely focused on improved and faster communication. The pandemic compelled lawyers to adopt remote working utilities without demur, and a happy – or not, perspectives may differ – consequence was a willingness to adapt and deploy new utilities.

The days of lawyers resisting change – this author and some of his colleagues loudly protested being handed his first BlackBerry along the lines that such things were best left as fruit; none of us got very far with that protest, and the author equally owns to adopting, implementing and enjoying the benefits of remote working protocols years before the pandemic struck – are long gone.

Nonetheless, pre-AI there were UI and UX changes to practice, with little meaningful change to the quantum and quality of human engagement required across mandates. We used technology for limited and clearly defined purposes and, accepting rubbish in/rubbish out and assuming bona fides relied on the output. Illustratively, word processing documents, running redlines, .xlsx formulae and VB applications, sending email and IM messages and, frankly, all the things that were and continue to be taken for granted as being performed better by technology in an infinitesimal fraction of the time it takes a human to complete the same task.

AI has manifestly changed that paradigm. From quick precedent summaries and trend analysis to ostensibly well written drafts, technology now seems poised to deliver output which, pre-AI, we relied on humans for. It is trite but no less true – that is the opportunity and the challenge.

At present, AI application to inter alia M&A transaction execution broadly relates to: reviewing data to identify synergies, risks and connections; presenting results and summaries in the format required; originating draft transaction documents; tracking connections across the dataset reviewed and the draft documents, as well as draft consistency and completeness; and generally assisting with and facilitating transaction management. Of course, there are utilities that offer more, but this function set seems most prevalent in India as of date.

All of this is quite literally wonderful. We continue to be amazed at receiving cogent and seemingly complete summaries and reports in an instant. Originating a similarly apparently excellent draft from a termsheet in practically no time is a delight that has yet to wear out. And this apparent excellence is the rub.

Self-professed Luddite tendencies aside, each of us has reviewed a data summary which is correct of itself, but has not addressed the nuance of the issue summarised, and so missed the substance of the risk. This applies equally to summaries of conversations in meetings. We often find that a connection is properly drawn and recommendation properly made, which recommendation, if implemented as is, smacks of transaction impracticability.

In many ways AI functions as an assistant, maybe even a paralegal, with a remarkably quick turnaround time. The experiential difference is that a senior lawyer reviews work product from a human paralegal with a mindset materially different from the mindset that we bring to work product originated by technology. We are used to, and consequently biased towards, accepting the results of technology usage as absolutely ideal and perfect.

In some instances, AI functionality and usage is capable of being treated as such, but these instances only build on existing technology solutions, and facilitate and expedite implementation, illustratively and simplistically, using AI to set up a calendar appointment following from an email message. They are of course game changing and add to connectivity and efficiency, but are not really experienced lawyer value-added functions.

In others, the more value added and so professedly disruptive applications, AI is an exceptional utility and game changer, but the output needs human review and improvement.

So, we must implement practice methodologies which leverage the benefits of AI and overcome the inherent human tendency to accept the result of technology use as being 100% correct and better than human. This is exceptionally important for junior practitioners – their comfort with technology and flexibility is a skill and advantage that must be nurtured and cultivated. Equally, in the present meta of AI, it is crucial that they double down on getting to grips with detail and nuance, which is often missed by these utilities. CLE must additionally focus on this.

So, as things stand, lawyers must adapt and change to address the opportunities made available, and the risks posed, by AI without losing pre-AI skill sets; at least, until AI achieves the ubiquity and reliability of the Eyeball Mk1 (ideally backed by service provider warranty and endorsed by professional liability insurers).

This is particularly relevant where AI is applied to M&A, given that each transaction is unique and requires bespoke solutions. For these matters the value proposition that AI holds is that of an exceptional assistant, but only an assistant.

That this will change is clear. Practice is moving from precedent to prediction, and siloed expertise to disseminated knowledge, and AI will evolve to accelerate this process. The benchmark for client engagement and the expectations that clients have from lawyers is higher than ever and AI, used by clients and by lawyers, will drive these higher still.

We will be able to more easily give in to our bias in favour of technology. It may be that in a short while we will receive AI output that maps to value-added professional advice. When that happens, and it may be a reality much sooner than expected, we will need to evolve and apply appropriate thought paradigms to practice – even to M&A.

The need to habitualise innovation has never been more pronounced.