Managing corporate income taxes effectively is important since these represent the second or third biggest expense for many enterprises. However attractive minimizing them might be, following all applicable tax laws and regulations is equally as important because failure to pay what the laws require can result in significant monetary fines and, potentially, reputational risk. Since there can be some degree of interpretation in what constitutes “net income,” tax departments are constantly challenged with trade-offs in tax accounting in deciding how aggressive they should be in making decisions in the grey areas of the laws. Too conservative an approach limits economic returns to the enterprise and therefore reduces its competitiveness and viability. Too creative an approach has potentially negative monetary consequences for the enterprise and professional ones for those involved. Adding to the complexity of the task, tax laws and regulations are forever changing in democracies in response to lobbying from one group or another seeking to raise or lower who is paying the taxes. Fortunately, AI in all of its forms is poised to disrupt managing corporate income taxes for the better. AI won’t put tax professionals out of work—to the contrary, it’s going to make them more valuable.
Managing income taxes is a challenge because there is a fundamental disconnect between business and tax accounting, with the former determined by generally accepted accounting principles and the latter calculated by the tax code. If you have spent any time studying the topic, you may conclude that any connection between the two is purely coincidental. There also is a difference in the concept of time periods. Whereas financial and managerial accounting exists in a largely continuous time span with periodic measurements, income tax accounting is done on a frame-by-frame, year-by-year basis. Think basketball versus chess. Consequently, there are myriad adjustments made to tax accounts to reflect year-to-year changes in laws and interpretations of the tax code. Tax codes themselves are governed by different concepts. Non-practitioners think of taxes as something that must be paid out of their pocket, while tax policy people think in terms of what they’re letting you keep. (This is the essential thought underlying the concept of tax expenditures.)
Note that when it comes to tax law, there are meaningful differences between countries in the tax concepts they use and how they are applied. The contribution to public revenues also differs between countries. For example, income taxes have traditionally been a more important revenue source in the United States than in continental Europe and other places. Historically, the latter have had a meaningful lack of compliance, hindering the government’s ability to obtain needed revenue. For multiple structural reasons, those countries have relied more on indirect levies such as value-added taxes (VAT) because these have been easier to administer and determine with a high degree of reliability. The difference has narrowed a bit over the past decades as increased digitization of accounting systems has made income tax evasion more difficult.
The income tax provision process is a core responsibility of tax departments and an essential part of the accounting close process. This procedure estimates the amount of income tax an enterprise will have to pay tax authorities in the jurisdictions in which it operates. Tax accountants derive the number by adjusting the reported net income with a variety of permanent differences, such as expenses that are not deductible and temporary differences—for example, using allowable accelerated depreciation for tax purposes and straight-line depreciation for financial reporting. Tax provision is one area where the need for digitization has grown over the past decade as countries impose new regulations designed to increase revenues.
For example, base erosion and profit shifting is a set of standards and support established by the Organization for Economic Cooperation and Development (better known as the
Using a dedicated tax application enables tax and accounting departments to operate more effectively. It allows the entire department to manage a consistent set of tax-sensitive data in a controlled process that promotes accuracy and auditability. Software promotes data integrity and efficiency by managing provision as an end-to-end process that takes numbers directly from authoritative source systems to construct tax financial statements, calculates taxes owed and keeps track of cumulative amounts and other balance sheet items related to taxes. Having tax data and tax calculations that are immediately traceable, reproducible and permanently accessible provides executives with greater certainty and reduces the risk of noncompliance as well as attendant costs and reputation issues. Maintaining an accurate and consistent tax data store enables enterprises and tax departments to better execute tax planning, provisioning and compliance.
The application of artificial intelligence (AI), generative AI (GenAI) and agents to tax provision software will make it even more useful in streamlining operations and giving tax professionals more time to apply their experience and expertise in dealing with the broad grey areas of tax law. Some of the mechanics of tax provision are straightforward in concept but complex in scope. A simple example is the rate-reconciliations process companies use to describe the difference between a statutory tax rate (the nominal, headline percentage) and the effective tax rate. The latter is the percentage accrued once things like deferrals and time differences are considered. The rate-reconciliation process details the rate drivers, those items that move the rate up or down relative to the statutory measure. These might reflect permanent differences or state tax expenses, as well as differences in foreign and domestic tax rates. Existing tax software can handle this better than humans because it automates the calculations, standardizes their drivers and controls all assumptions at a central point. Reviewing the rate-reconciliation calculations is faster and more accurate than a manual spreadsheet-supported process. AI will take this process a step further, speeding it up while making it more reliable.
The ability to use predictive AI and GenAI in managing taxes will enable departments to consider a wider range of approaches in their tax treatments in a more consistent and objective fashion. Transfer pricing is one obvious area where optimizing routines will make it possible to find the best mix of prices, their justification and risks, while providing a narrative of high-level and detailed considerations behind various options. Tax-related AI use cases initially will be confined to areas where vast amounts of number crunching currently create a time barrier to their usefulness. Assessments that now require hours or days to research, explore, assess and document will be available in minutes.
At the same time, more interpretive applications of the technology at the enterprise level are still somewhere in the future. It’s unlikely that a useful corporate tax-advisor-in-a-box will be generally available anytime soon, but tax advisors themselves will have considerably greater ability to apply purpose-built language models in assessing and creating tax strategies for corporations, especially those operating in multiple countries and regions.
Tax management processes have become more effective and productive over the past decade-plus as the software necessary to manage the complexities of tax provision and planning has improved. Along with this, the role of the tax department executive has become more strategic. Here’s one indirect measurement of the shift: 15 years ago, I couldn’t find a single senior vice president of tax on LinkedIn; today there are over a thousand. AI-embedded tax provision, planning and compliance software is poised to substantially improve the productivity and effectiveness of the tax department. Embedding AI in all of its forms will further increase the business value of enterprise tax software. I strongly recommend that enterprises that still use spreadsheets or internally developed software to support an even moderately complex tax provision process evaluate a dedicated application for this purpose.
Regards,
Robert Kugel