The value of human expertise over process-driven outcomes

For a year or so, there has been much talk about how AI will not replace accountants, but free them of mundane and monotonous tasks and allow them to spend more time with clients. However, worries about job losses and role replacement persist. Joe Pickard speaks to TGS Global CEO Andrew Menzies about accountants’ concerns about AI and its impact

The Accountant: What are the main, specific concerns that your members are having about AI?

Andrew Menzies: At present it is more a general unease. They can see that compliance work of annual accounts and tax return preparation will be able to be done more cheaply, quickly and easily using AI in the near future. They are uncertain whether they will be able to capitalise on it, or merely be overtaken by online services that avoid them altogether.

AI is already being used by clients to give them answers to problems, but at present they tend to ask our members to get a reality check. The fear is that this will stop.

Our members accept that lower-level jobs will be taken by AI, but then they wonder how they will train the next generation of senior accountants to take over from them at the higher level of thinking. Indeed, at the current rate of development of AI, will there be a need for senior people anyway?

On a more granular level, they have concerns about how to introduce AI into their workflows, and whether, once they rely on AI in that way, the big AI providers will increase their pricing once they have enough businesses hooked. And they worry about liability. AI, like humans, gets it wrong sometimes, and what will that mean for them? Conversely, just as we see AI being used to bring a vast number of cases in the employment tribunals today, surely it is only a matter of time before AI is used similarly against them. The cost and reputational damage of a claim, even if it is won, can be very high.

Our members can see that a new world is emerging, but they cannot yet sense the direction it will take them.

TA: Are these concerns universal? Do they vary between types of firm, geography and so on?

AM: Broadly, the concerns are universal. Firms more focused on compliance work feel most exposed and are seeking new strategies. Some are looking at how to capitalise on the use of AI within their workflows, others see this as the right time to exit compliance and move to a more advisory and consulting framework. Audit firms are keen to use AI tools to simplify audits and increase the certainty of the results of audit testing.

There are some geographic differences. In Europe, as a developed and highly regulated region, AI feels like a greater focus of interest and simultaneously, concern. In the Far East, technology is seen very positively, and there great efforts are being made to embrace AI, but similar concerns always remain.

North America has greatly embraced AI, while Latin America, again, wishes to embrace it but is fearful of its US owners.

The underlying theme seems to be about trust. Can a firm trust AI answers or results? Can a firm trust the US-owned companies that provide the AI? Our members are experienced and professional, but they also come from many nations, and most would be more comfortable if the AI they use were controlled – if it can be controlled – within their own known territory.

Another universal theme seems to be about direction. When desktop computers invaded the accounting profession, it was clear what they could do and how they would do it. Of course, over the years they have improved beyond recognition, but the technology remained the same. When the people building ‘frontier models’ tell you themselves that they do not understand what their AI is doing, it becomes a greater concern.

And finally, a theme that seems to be growing is that if, in their next iteration, these AI models can develop economic models of their own, financial modelling beyond our comprehension, analysing data in markets in a manner so beyond our understanding that we cannot compete with it, what will this do the very fabric of our financial markets and banking structures?

Of course, this is a bigger question, but one that does get talked about. One commentator recently suggested that even if the AI were to explain what it was doing, it would be a little like us trying to explain bond yields to a cow. We simply could not understand it.

TA: How are you handling member worries, are you finding tackling it with a light-hearted approach is working to dispel concerns?

AM: At present we hope that the main change will be no more than getting rid of the grunt work of accounting. What this will do to training and future understanding is a more difficult problem. However, we cannot really dispel concerns – we all have them and they are real.

The light-hearted approach is intended to get people to engage with thinking about it, but not to provide reassurance. After all, what reassurance can anyone give about the future?

Our intention is to ensure that those who are not considering it, do so. Also, as the theme underlying the article tries to explain, the entire world evolves and changes, and we cannot stop it. Often we cannot see how things are changing except in the rear-view mirror, but ignoring change is a recipe for being left behind. Making the wrong decisions now will most likely lead firms to suffer in the future, but we are in no stronger position to advise our members on the right decisions than they are themselves.

We encourage them to consider their own strategies and not to ignore the changes afoot. Each TGS Global member is independent and it is not for us to direct their behaviour or strategies, only to help them to see what we observe.

Our main activity is to provide expertise at TGS Global conferences, where we invite experts to speak and debate with our members on many subjects. AI seems to be the most engaged of these discussions.

TA: What guidance are your members looking for when it comes to implementing AI?

AM: Our members are not seeking our guidance; they are looking for understanding. It is dangerous to suggest that anyone has answers to what should be done, and what the future holds. We seek to provide them with some of that understanding through expert speakers.

Our view as a network is that there are those who have a better understanding of the technology than others, and who may be able to guess what the future may hold, but even the very best of them admit that they do not really know.

Our members might not be looking for it but they are certainly getting continual and updated guidance on authorised use and presentation of the TGS Global brand. Our brand is registered and protected so AI transformations are a concern for us.

We are in the process of redesigning our brand manual with AI in mind, and we are seeing a wild variation of brand interpretations from the various tools our member firms use to create their communications.

It is difficult to keep an eye on and to give up-to-date advice when the tools themselves are evolving so quickly. One week the AI tools might create something acceptable and then a couple of weeks later the same prompts can often lead to substantially off-brand results.

We try to reduce this negative AI impact on our marketplace portrayal with a combination of checklists, prompts and information webinars about what is authorised, what is encouraged and what best practice is. Our members are looking for this kind of support, and they find it useful to have a whole network of humans in the loop.

TA: What impact is AI having on employment at your firms?

AM: As mid-tier firms, most of our members remain more people-oriented than process-oriented. Even our larger firms place higher value on human expertise than process-driven outcomes. This is what their clients currently want and expect.

At present, we have not yet seen a dramatic fall-off in the pace of recruitment, but we know it is happening in the industry. In some countries, student intakes may have reduced marginally but that has not, so far, been put down to AI, but rather on reduced staff churn – which may, in turn be the result of staff fears about AI.

Although most firms wish to stay ahead of the technology curve, accountancy does not seem to wish to change until change becomes the only option – accountants are broadly conservative in their outlook. It is worth remembering that, except in the largest firms – who are not members of TGS Global – the firms are led by senior professionals who will keep their eyes open but will drive in straight lines as far as is possible.

In fact, it is becoming clearer to us that, possibly, whereas historically the Big Four trained most accountants, many of whom would then leave to bring their expertise to mid-tier firms, the future might bring the reverse, where most training is done by mid-tier firms whose clients seek more personal service and are prepared to pay for it. These staff will then take their skills to the larger firms whose training will become ever more limited. But this is all crystal ball-gazing.