Artificial Intelligence in Fund Board Governance
By Tarne Bevan, Business Manager Europe, M&G Investments
Many of us will recall the hype over recent years around new technologies such as blockchain/ distributed ledger technology (DLT) and more recently tokenisation, which were (and are still) expected to fundamentally change the fund management industry by reducing costs for investors, improving margins for managers and improving the security and scalability of operations. Some may even recall the Dot Com boom (and bust) which resulted in web-based technologies and mobile devices fundamentally altering business models, processes and engagement channels with investors. This hype drove high return on investment expectations for investment firms and paralleled concerns over job losses. The pessimists (or perhaps realists) put much of it down to excessive hyperbole for an industry with somewhat of a luddite legacy. The reality has been somewhere in between.
The latest hype around Artificial Intelligence (AI) raises similar expectations. Is this time different? AI has been around for several decades. It is already well embedded in supporting investment processes, risk modelling, robo-advice and virtual call centre assistants to name a few contexts. So why have we not seen materially expanded margins and plenty of job losses? The reality is that there are various forms of AI with most augmenting our capabilities and resources, not replacing them.
There are those that rely on defined parameters and rules but not historical data, and these can process significantly faster than humans – think of scenario modelling. Imperfect data can be a challenge to their efficacy. And let’s face it, our industry is riddled with imperfect data, but much of the ‘return on investment’ was really just digitising existing processes to make them faster and less prone to error.
There are those able to rely on historical data to make predictions and handle complex classifications – think self-driving. These models still struggle with outlying data and concepts. However much of the AI we have been hearing about in recent years, including Generative AI, falls into this category.
The more futuristic types aim to replicate human-like thinking, judgement, reasoning and motives. And beyond these we start entering the feared category of AI that is destined to overtake human intelligence altogether. We are a long way from these realities and the increasing focus from governments and regulatory bodies are likely to slow their evolution, for the better.
So, what’s different this time, and is all the hype realistic? To the former, Generative AI tools such as ChatGPT have changed the playing field in ‘usability’ thereby enabling a wider population of non-technical users to make easy use of its capabilities. This leads to a material opportunity to reduce administrative activities and analyses for non-technical staff such as summarising topics, report-writing, email drafting and, importantly for the world of governance; minutes drafting, resolutions drafting, summarising board packs and supporting effectiveness reviews. Regarding the hype – is this going to wipe out huge numbers of jobs? It will likely lead to some headcount reduction in some industries and job types more than others. However, the reality is it will likely augment many, much as past innovations have.
Consider the preparation of minutes as an example. Since COVID we’ve learned to be comfortable with recording meetings, so audio and text transcripts have been one improvement to support the job of company secretaries. Making use of these for multi-hour meetings, with the help of Generative AI tools, can cut down minutes preparation to a small percentage of that time. Of course, they will still need to be reviewed for some fine-tuning of accuracy. And we all know that transcripts can pick up a lot of noise in meetings such as vocal accents, multiple people speaking at once, jumping back and forth around topics. Improving meeting discipline to reduce some of this noise, along with thinking about how to better structure meeting materials with agendas and decisions/actions, wouldn’t be a bad thing. And by doing so, the benefit of Generative AI tools will only improve. Thereby enabling the governance process to be enacted with less effort and shorter times to execute on decisions and actions of the boards and committees, freeing up time for higher value-add activities.
Tarne is the Business Manager for Europe at M&G Investments, based in Luxembourg. She is responsible for a number of commercial and oversight activities related to the international distribution and management of M&G’s funds and investment capabilities. She also manages the Company Secretary function which supports one of Europe’s largest and most complex ManCo’s and a range of fund boards.
