Many tasks previously carried out by juniors have now been taken over by artificial intelligence (AI) and, in the last two years, the Big Four accountancy firms have reduced recruitment for junior roles. However, with offshoring and general cost reductions factored in, it’s unclear how much the growth of AI contributed to the decision. 

Meanwhile, ICAEW’s Evolution of Mid-Tier Accountancy Firms research reveals that 68% of those surveyed expect AI to reduce the need for some early-career roles. It also showed that advance of AI is creating uncertainty around what skills are going to be needed in the future. 

We asked early-career professionals for their views on some of the key issues. 

Building foundations

One of the concerns expressed by senior accountants is that the increasing automation of routine tasks may mean that trainees miss out on some of the foundational skills and knowledge that come from doing them manually. 

However, Chris MacDonald, an ICAEW student and junior accountant at DG Accountancy Services, believes that such automation will make the profession more appealing. “Sure, it's important to learn the building blocks and cut your teeth on some tedious bank reconciliations and bookkeeping, but there is only so much of this you can do and stay motivated. I think AI, when used well, can get new recruits on the meatier and more interesting stuff quicker,” he says.

Erin James agrees. The accounting trainee at PKF Francis Clark, who is in the final year of her ACA qualification, says: “I understand these concerns. However, I think some of the tasks AI has taken over aren’t ones I would necessarily have learnt from, for example, exporting a bank statement to Excel instead of having to type it out. I will still be analysing the bank statement as I would previously – AI has just taken away the non-value-adding activity.” 

Matthew Laurenti, Assistant Manager, Business Services, at PKF Smith Cooper, believes that employers need to take responsibility for ensuring trainees still gain a good grasp of the basics. “As a manager and reviewer, it is easy to see when a trainee has relied too heavily on AI as when questioned they’re unable to provide answers confidently. Questioning the results, even when they’re right, ensures that trainees have not just come to the correct conclusion, but also that they understand the logic and reasoning behind it.”

The risk of too much early responsibility

What does concern MacDonald is the need for adequate training and the risk of junior staff being tasked with AI oversight too early on. “I think this could be a potential banana skin. AI will only be as good as the user, and if what you're feeding into the AI is bunk, then the outputs are going to be rubbish. Then, if you have more junior recruits checking the outputs, there is the potential for something incorrect to be given the green light.” 

AI is also likely to accelerate the trend towards more advisory work, interpretation of data for clients and relationship building. Will junior staff be comfortable with this level of client interaction early in their careers? 

For MacDonald, who previously worked in recruitment, it’s not a problem. But he feels some younger accountants may struggle initially. “Older generations grew up in a world where picking up the phone was the default. With millennials, and even more so with Gen Z, their path of least resistance is WhatsApp and a phone call signals to them something is wrong. But Gen Z are very attuned to fast changes in trends, plus there’s a real hunger for genuine connections and authenticity now, so I think in the long term the drive towards advisory services will actually work in newer recruits' favour.”

James agrees that communicating with clients can be daunting at first. “Luckily, we have a lot of support from our managers. I think a lot of people want to get to a stage in their career where they have more client contact, and with AI taking over some of the administrative tasks we have more opportunities to learn about advisory work and attend client meetings with managers,” she says. 

“There was a time when I felt slightly overwhelmed in a client meeting and didn’t feel like I could answer all their questions, but looking back I think this was a good learning opportunity. Now before a meeting, I try to anticipate any questions the client might have and seek any additional help from my manager beforehand, so I can be confident in responding.”

Looking to the future

Everything that AI produces must be reviewed by a real accountant, says Laurenti, citing an experiment where he asked an AI tool to analyse payroll data.  “I wanted annual totals for certain members of staff from several monthly data sets. The AI tool was not even able to sum correctly. While it is constantly improving, there is still a lot of work required. AI is also only as good as the person using it. If the prompt is badly written, the result will more than likely not be accurate or the desired outcome.”

James is excited by the movement towards AI. “It will help keep my role varied and mean that I can spend more time gaining technical knowledge and developing skills in areas that add value.”

And MacDonald isn’t worried about job security. “I think the talk of AI taking all of our jobs is largely doom-mongering. I think it’s very much a case of AI being a blessing or a curse depending on how you use it. If used diligently, it can take care of some very dull tasks and leave you to spend time on the things that actually make you get up in the morning.”