8 minute read 18 Nov 2019
Man talking telephone checking business data computer

How AI looks to upend the tax function’s 80/20 rule

By EY Global

Ernst & Young Global Ltd.

8 minute read 18 Nov 2019
Related topics AI Tax Tax function operations

The powerful combination of artificial intelligence and human expertise opens up a huge range of possibilities for the tax function.

Artificial intelligence (AI), one of the key technologies of the fourth industrial revolution, is transforming how businesses operate. AI-based systems are capable of ingesting information and instructions, learning from interactions with data and humans, making decisions and responding to new situations and questions in a precise and logical way.

This capability is ushering in a new era of visibility into transactional data and the ability of companies to unlock value, help manage risk, improve efficiency and provide critical business insight. The tax function is no exception.

Armies of tax accountants still laboriously trawl through corporate accounts to determine tax implications. It’s meticulous work that requires the knowledge and experience of a highly skilled tax professional. But a single machine using AI can now do much of this work in just seconds with improved accuracy and consistency, freeing up tax teams to provide more valuable contributions to their organizations.

According to Martin Fiore, Managing Partner, EY Northeast Region Tax practice at Ernst & Young LLP, 80% of the tax function’s time is spent gathering data and 20% analyzing it. “AI allows us to almost invert that, creating a massive gain in value,” he says.

Slashing manual compliance time

For example, AI is transforming the classification and analysis of P&L data for tax and reporting purposes. Automated ledger review tools already use machine learning, a branch of AI, to transform the classification of transaction-level data for a variety of tax and reporting purposes. 

With traditional processes, analyzing and classifying financial data to determine its tax treatment can be a manual and time-consuming exercise with no guarantee of consistent results. This approach is unsustainable, given the increasing volume of data to be analyzed, cost pressures within finance and tax functions, accelerating reporting and compliance initiatives, and the growing digital demands of tax administrations.

The average time for a human to analyze 10,000 lines of P&L data is 75 hours, but AI tools need just one minute, thereby freeing up scarce resources, and improving consistency and confidence in the results. “To put this in perspective,” says Dr. Harvey Lewis, Chief Data Scientist for EY’s Tax Technology and Transformation team in the UK, “just a month of transactions for a larger company might create anywhere from 16,000 to 25,000 lines of data.”

AI is helping businesses with all types of compliance, from the deductibility of expenditure for corporation tax purposes to capital costs allowances and recovery of VAT. Charles Brayne, EY UK Tax Technology and Transformation Team Leader and the EY UK Tax Innovation Leader, cites the example of a UK-headquartered FTSE 100 client seeking to enhance compliance outcomes while reducing time spent meeting employment tax requirements. “We worked closely with the client to provide an approach, making use of both robotics and AI to introduce consistency and rigor to employment tax decisions, enabling the company to radically reduce time spent on repetitive, labor-intensive basic compliance,” he says. “The solution not only helped the client to improve the efficiency of their internal processes, but also to respond to the increasing digitization of major tax authorities who are using technology that increasingly includes their own AI.”

Monitoring change and answering questions

EY teams have also deployed other AI technology that can help companies monitor tax-relevant events around the world. One tool, EY Salient, allows a company to keep on top of developments in a way that a single human individual could never do.

“You could spend your entire life just browsing news and papers that may or may not be relevant,” says Lewis. “And how long would it take to read 300, 400 or more tax authority websites, work out which changes have just been announced on those websites, then filter it all to determine which tax and which jurisdictions they’re talking about?” AI can increasingly do all this and more.

This is powerful technology at both the global or local level. For example, says Lewis, the responsibility for remaining compliant with local tax codes often falls to local country tax managers, who are expected to monitor tax developments on top of all their other day-to-day activities. As a result, too often updates fall between the cracks. Similar problems occur when companies learn of a substantial change in one jurisdiction, he says, because – with limited resources – there can be a tendency to focus exclusively on that jurisdiction to be sure relevant processes and data remain compliant, while potentially missing another change elsewhere, which might have a material impact on their business.

We’ve helped a client deploy a virtual assistant — in this case, a chatbot platform — to give tax advice to their finance teams.
Charles Brayne
EY UK Tax Technology and Transformation Team Leader and the EY UK Tax Innovation Leader

“Keeping an eye on this from a global perspective is quite a challenge,” says Lewis, “but AI can provide both global and local information almost on a real-time basis, helping companies to identify the developments that are material to their business and giving them early warning as well as greater knowledge, oversight, insight and risk management at their fingertips.”

AI-powered technology is also being used to handle many of the routine, repetitive questions that tax teams get from business and finance stakeholders. “We’ve helped a client deploy a virtual assistant — in this case, a chatbot platform — to give tax advice to their finance teams around how they code and classify transactions from an indirect tax perspective,” says Brayne. “We worked with the client to generate the content and trained the chatbot appropriately to give the right answers.”

Stakeholders can now simply access the information by asking the chatbot a question. It’s on demand, so they don’t have to wait until the tax team is free to provide an answer. The chatbot can also ask clarifying questions when needed and if there are answers the chatbot doesn’t know, it will flag it to the tax teams. “Ultimately, end users are getting the right answer, quickly, and it’s being done by a machine instead of a human in the tax function,” says Brayne.

For Channing Flynn, EY Global Tax Technology Sector Leader, there is much to get excited about as we look further ahead. “In the future, AI will allow unprecedented reach and insight into data that will transform how transfer pricing comparables are done,” he says. “Currently, transfer pricing is more art than science and a major challenge for both taxpayers and governments alike. AI will also radically change controversy and planning – both time-consuming and complex areas. AI will allow companies to manage risk and do predictive analyses far more quickly and powerfully than is possible today. Companies will be able to gain rapid insight into case law and procedural challenges and all that’s available in the public domain to help inform a tax strategy or to update tax compliance procedures for new law.”

With the rapid developments in machine learning, data mining and cognitive computing, the next decade promises to see huge leaps forward in AI. “As powerful as AI is today, it is still in its infancy,” says Flynn.

People and machines: better together?

While the potential benefits of AI are huge, there are some well-publicized fears to overcome. Foremost among these is that AI will replace humans, but Flynn is keen to reassure that an alternative reality is more likely. “AI will enable tax professionals to work better, smarter and faster, which will mean that they can spend more of their time providing more valuable activities for the organization, from managing risk to partnering with the business on its strategic projects and priorities,” he says.

Moreover, AI will create new roles that never existed before, such as solution architects who need to know both technology and tax, and enhance the ability of the tax profession to attract and retain the best and the brightest, according to Fiore. “The higher-level, higher-tech work that AI allows will encourage the best and brightest to enter the profession, it will help them learn faster and it will provide greater career satisfaction,” he says.

It is also important to note that there are some things AI cannot do, including providing human skepticism, interpretation and judgment. “AI will not do away with the need for human experience and insight,” says Flynn. “Highly qualified and skilled tax professionals with sound judgmental minds will still be needed to make good decisions about what an AI-powered machine concludes or recommends.” 

 

It’s not about replacing humans, it’s about technology augmenting the role that humans play.
Dr. Harvey Lewis
EY UK&I Chief Data Scientist – Tax Technology and Transformation

AI applications need to be trained by people who know tax. “It’s not about replacing humans, it’s about technology augmenting the role that humans play,” says Lewis. “We can’t get a machine to be absolutely perfect in the same way that we can’t get humans to be absolutely perfect. Both make mistakes, but they tend to make different kinds of mistakes. The collaboration enables clients to do things they’ve simply not been able to do before.”

Ultimately, just as AI is in its infancy, most organizations are at the beginning of their journey with the technology. “What should be holding companies back is a lack of talent, but it’s more a lack of understanding of what’s possible,” says Lewis.

Action points

  1. Identify what the tax function’s digital road map looks like today and what stakeholders would like its digital journey to look like.
  2. Identify where AI technology can add value.
  3. Use alliances to help identify appropriate tools, application interfaces and learning data sets.
  4. Acquire tax talent — whether internal or external — with the technology expertise to support an AI initiative.
  5. Design and implement a proof concept to experience and assess the future potential.

Summary

Integrating AI into the tax function can help improve efficiency and satisfy the growing digital demands of tax administrations.

About this article

By EY Global

Ernst & Young Global Ltd.

Related topics AI Tax Tax function operations