LONDON – Across the board, AI adoption is on the rise. A recent EY survey of US CEOs and business leaders found that most senior enterprise decisionmakers (85%) classify themselves as artificial intelligence optimists, while the financial services industry in particular is benefiting significantly from an uptick in AI adoption.

One area of financial services quietly undergoing a revolution in AI and automation is tax and accounting. To gain a better understanding of how AI is being used in tax today, we sat down with Charles Brayne and Harvey Lewis from EY. Charles is EY’s UK Chief Tax Innovation Officer and Partner and has a dual role. On the one hand, he works with clients to help them adopt new tax technologies and on the other, he oversees the implementation of AI technologies within EY’s own tax business. Meanwhile, as EY’s UK Chief Tax Data Scientist, Harvey works directly with tax and law professionals to create and deliver new AI tools and applications, as well as provide strategic oversight for their automation projects.

Why is AI a focus for tax?

C: Whilst tax is enormously complicated in so many ways, it’s also very systematic and logical. We’ve found that a lot of what we do is enormously susceptible to being performed in a very different way using various branches of AI, such as pattern-based machine learning, Natural Language Processing (NLP) and Natural Language Generation (NLG), all of which we’re investing in heavily. In all likelihood what we do as tax professionals will be fundamentally transformed in the next 3-5 years.

In what way?

C: A lot of the routine work in tax is around compliance, data cleansing and data processing; being able to identify risks and anomalies, and produce returns. We are now starting to use technology to deliver this work in a fundamentally different way. People are going to be using a range of technologies to produce those outputs. You could see fewer people doing tax compliance and accounting in future because so much of it will be driven by technology, or at the very least, those people will have more time to focus on the outputs of that work and what it means for the business.

What do you see are the benefits, Harvey?

H: For me, there are two primary benefits for AI in tax. The first is that it allows you to deliver old services in new ways. If you think about some of the conventional tax or law-related services that organisations provide, AI can make them much more efficient by orders of magnitude.

The second is that entirely new services are possible when you apply these technologies. We can look at the data and tasks that we perform and apply AI to enable something very different, so whether that’s screening new information sources for changes in tax legislation or helping tax professionals complete interviews with engineers claiming R&D tax credits. There’s a wide range of applications for these technologies that allow us to go much wider and deeper into these services.


Related: Joining the dots – overcoming business problems with AI


Where does all of this fit into the future of tax?

C: This isn’t just about tax professionals, but tax authorities as well. We are starting to see fiscal authorities invest in these technologies to analyse large volumes of data, make sense of it, and to understand where the tax revenue may be falling short. The UK as well as fiscal authorities around the world are investing very heavily in technology to support new ways of implementing enforcement and collection mechanisms, using more data sources than ever before. So actually, the technology is starting to change the very nature of tax compliance regimes around the world.

The other aspect is that forward-thinking finance leaders are recognising that what we used to think of as tax compliance processes are really finance processes at heart, because a lot of this classification and categorisation work can be done by a finance professional using the right technology. This is leading to quite a profound reshaping of the relationship between the tax and finance function, including a fresh look at where tax professionals should be focusing their attention in order to add the most value.


Related: Addressing the implications of AI for individuals seeking payday loans


In the next twelve months, what should companies really be focusing on to facilitate the growth of AI in tax?

H: The challenge here isn’t about starting – it’s about sustaining. I think it’s relatively easy to start: you can define a set of use cases, you can find some technology, and you can start the process. It’s much harder to sustain that and scale it across the business. You will have to start with the right use cases – those that really add value. For the clients we work with, picking on a “difficult” use case such as the way the tax department delivers compliance, then demonstrating that an AI-driven approach can be scaled globally has been very powerful. The key is to have sustainability in mind from the outset. With cloud and server-based infrastructure, you can scale very rapidly, so the way in which you implement the technology is really important for the way you can sustain growth across the business.


Catch Charles and Harvey’s keynote, ‘Lessons from implementing AI in a people business’, at 09:30 on the Implement stage at The AI Summit London, June 13. Find out more