Mid-tier and small accountancy firms need to create space to fail in their efforts to trial, test and iterate the implementation of artificial intelligence (AI) into their audit processes, says Andrew Moyser, partner responsible for AI and data analytics implementation at MHA.

AI has become the byword of business in a short space of time. However, most mid-tier and small accountancy firms have yet to fully implement AI tools into their audit processes despite the promised efficiencies and time saved.

With technology moving at such a pace, Moyser says efficiencies will inevitably emerge as new technology is embedded in their ways of working. For now, though, his focus is on quality of data. “Efficiencies come later once people really understand the technology and it is embedded within the methodology,” Moyser told delegates at ICAEW’s one-day annual conference earlier this month. 

“Those efficiencies will come tenfold once we’ve figured out exactly how we’re using and capturing best evidence. You’re better off making decisions today about improving the quality and accuracy of the data that you’re getting.”

“In testing and retesting the quality of data input, the ability to fail should be front and centre in the evolving audit sector,” says Moyser, who leads the firm’s audit innovation project and is responsible for bringing data analytics and AI into its audit procedures.

The added benefit of this, he says, is that junior auditors will be attracted to and able to participate in the evolution of audit. “If you start bringing in technology that can make the sector more interesting, that really is something that the more junior team members really enjoy and can certainly seek the benefits of.

“Like any bits of technology, you’re better off trying things and failing than not trying at all. When you’re looking at technologies in the market, sometimes you don’t always have the answer. You’re better off trialling it and learning something. And even if you learn that it doesn’t work, that’s a really important lesson in the development of anyone’s audit practice.”

The Big Four accountancy firms have been investing heavily in advanced technologies over the past decade and have reached a point where they have proprietary data analytics tools and AI tools. But the majority of firms remain at the investigating stage.

Ian Pay, ICAEW’s Head of Data Analytics and Tech, says the ability to fail needs to be done in a safe and responsible way. “Failing on a live client engagement would not be good from a regulatory perspective, we really don’t want to be encouraging firms to do that.”

However, Pay says there are ways for firms to manage experimentation with AI tools, including trials on dummy or prior year data (where permission has been obtained to retain that information). “Firms can also run new AI/tech in parallel to existing audit processes, to allow for comparison of results or to fall back on those existing processes if the new tech does fail.” 

Pay also recommends testing new approaches prior to year end – even as early as after Q1 or the half-year – on smaller data sets, so it is possible to identify early and quickly if the new approach will work.

Partnering

Many firms are finding new ways of doing business by partnering with technology businesses to analyse what the technology can do and how it can be implemented into the audit process.

“We’re very much focused on partnering with other tech businesses, and utilising that technology and seeing how we can adapt that into our infrastructure and into our audit practice,” says Moyser, who ensures the firm’s audit quality as he reviews its high-profile clients and more complex audits.

“It’s about understanding what’s coming to the market and what’s new, and then understanding how we could potentially use that within our audit practice to either improve efficiency or improve quality.”

Getting data from clients has become easier, but some legacy systems still pose challenges to auditors. The ability to run tests on bigger sample sizes and analyse that data at a more granular level is a significant help to improving data quality, he says.

This learning curve is a two-way street, though, Moyser explains, because sometimes new tech providers don’t appreciate that some of the legacy software providers can’t just tap into an application processing interface. “The mid market generally is handcuffed by some of the software providers they use,” he says.

Pay stresses the importance of transparency when partnering with technology businesses: “It is important that the third party understand that regulators will want to know how the tech works, and this should form part of implementation discussions.”

The other interesting area that’s emerging is the ability to access more data at speed. “Even simple things like chatbots, or ChatGPT to some degree, are useful technologies because they allow a user to gain access to information very quickly. If you can use AI to look at things like your firm’s policies and procedures at speed, that then helps with liaising with clients and communications channels,” says Moyser.

He warns that creating efficiencies from technology won’t always result in reduced hours, “because staff then take that ‘saved’ time and focus on something else in more detail, or they’re looking into something that perhaps they’ve not looked into before. The efficiencies are not always clear because they are reinvested into other areas.”

For mid-tier and small firms, the cost of AI tools is a major factor. For MHA, Moyser said the firm will implement AI tools firm-wide when they “pass quality control, but also from a financial perspective”.

Ultimately, the evolution of audit is as much about people as it is about technology, he said, and it’s important to understand which employees will need access to AI tools. “Technology is half the battle. The other half is actually getting people to use it, understand it, work with it, and want to do that,” he says.

It’s also about future planning, Moyser says. Emerging technologies may not fit your firm currently, but what about the future?