Machine learning or people power?

Morgan Davies, director at Prime Accountants Group, looks at the pros and cons of artificial intelligence in the accountancy sector

Artificial intelligence (AI) is changing the way we work and live, and the world of accountancy is no different. The use of AI in accounting isn’t currently widespread but there is plenty of noise about it coming down the tracks, so we need to be aware of the opportunities.

Many people are saying AI will be a worst-case scenario for accountants, but it’s fair to say the accounting software market isn’t currently marketing any solutions. Where they are available, it is generally at the lower end of data input – for example, Xero has had a basic AI function which helps with the accuracy and speed of data input.

In theory, a well-developed AI solution will eventually replace bookkeeper jobs. However, if it will make the job quicker and easier, it also frees up the time and capacity of experienced practitioners to give even more advice to clients. This is where the big opportunity lies.

Once data is in place and reliable, we could use AI to do a lot more with that data, such as forecasting and benchmarking. We would then use the trusted human advisor to give more prompts and offer better and more accurate support. All those tasks sit with the humans right now, so software companies are missing a trick.

Over the next 10 to 15 years, I see opportunities to do more complex work more efficiently, proactively and cost-effectively using AI. Some of our industry’s services can be cost-prohibitive and this will help that – which, in turn, will improve the quality of advice to inform companies’ decision-making.

Currently, I don't think business owners trust AI, so we still need to interact and build relationships on a human level. In that way, I think it could make the bond between client and accountant even stronger. Maybe it will be a threat in 20 to 30 years’ time, when people do trust computers more than people! But that is something we will learn to deal with.

AI might be a faster, more comprehensive search engine than what we are used to, but it doesn't qualify whether the question you’re asking it is the right question to ask. That is one of the big skills you learn as advisors.

With a high percentage of questions asked by clients, we have to interpret them and work out what they are really asking. Unless there is a massive input of data about that person’s unique lifestyle and what’s important to them, how can AI ever understand their priorities? A personal advisor will pick that up via how they react to questions and rely on the personal relationship built up over time. The best advice comes from knowing the personalities, families and goals of the clients.

It also comes back to the core question, why do people use an accountant? Partly it might be because they know it’s a complex area and don’t want to upset HMRC, but it’s often because they want to improve themselves and they can't do that just by relying on IT. My most successful clients are the ones I speak to most often.

There is also the litigation risk. If you take advice from AI, it could get you in trouble and it will be down to accountants to bail people out and deal with the fallout with HMRC. Depending on the advice people get they will often have no recourse – if you get a book from the library, take its advice and lose out, you can’t sue the library!

However, the challenge of AI isn’t something to fear, it’s something to embrace and work with. Shop floor workers in factories have been at risk from progress for more than 200 years; it’s been something they’ve faced since day one. It is just our turn now!