Today, AI technology solutions are successfully tackling real-life problems all around us, every day. AI is proving to be impactful in solving medical conundrums, recommending movies (that I actually want to see), and may soon drive me to the office in an autonomous vehicle. It is exciting to witness firsthand how this powerful technology is transforming industries and improving lives.
"AI’s ultimate value in the field is crunching large volumes of data to produce greater insights and inform better tax decision making"
I am a tax lawyer by training and, like many tax professionals, I find myself constantly seeking creative ways to solve tax problems. So, how can AI technology be applied to tax? To help address this question, I decided to go back to school and enroll in the Massachusetts Institute of Technology (MIT) course “Artificial Intelligence: Implications for Business Strategy.” Here are my observations from that course.
I came away from this MIT course with the belief that AI has the potential to have an oversized impact on tax problem-solving. It is now clear to me that AI will affect taxation from quite a few different vantage points. There is plenty of research on the prospect when general AI solutions will equate to human natural intelligence, but that day has not yet arrived. Instead, as we near the end of 2018, there are tremendous opportunities for narrow AI tools that are uniquely developed and trained to solve specific problems.
Machine learning and natural language processing will undoubtedly have a profound impact on tax. Taxpayers will utilize natural language processing to extract and analyze unstructured data, and assist in tax determinations or produce tax returns. In addition, so much of core tax work today involves the mere classification of transactions, assets or people for tax purposes, that machine learning solutions are ripe to predict the associated tax treatment in this regard.
AI will also profoundly impact employment. Traditional tax professionals within corporate tax functions will be complemented with computer and data scientists adept at applying AI tools. These data workers will solve tax problems through AI/data analysis first, and specialized tax knowledge second. They will recognize that the issue of how AI will impact tax employment isn’t about machine vs. human tax professional. Instead, the relevant question will be how human tax professionals, with the assistance of AI tools, will make data-driven decisions that enable and enhance their productivity and excellence in solving tax problems.
I am also interested in how future AI tools will be developed to analyze the tax consequences of corporate actions. Most corporate tax departments are not large enough to address the plethora of daily transactions and issues that would benefit from a tax effect determination. In-house corporate tax professionals cannot possibly determine the entirety of tax consequences. In this regard, AI tools will be perfectly aligned to perform this valued function—in the moment, as issues arise that need immediate tax attention. For example, we may witness ubiquitous chatbots uniquely trained to answer tax problems. Just as we leverage digital assistants to answer general questions today, chatbots will be specifically configured to address technical tax issues.
However, the real promise of AI applied to tax problems lies beyond automation and the classification of transactions, assets or people. The real promise of AI in the tax field lies in achieving new insights not possible by human tax professionals. AI’s ultimate value in the field is crunching large volumes of data to produce greater insights and inform better tax decision making.
Rest assured, as taxpayers integrate AI solutions into their tax processes, government tax authorities would utilize AI tools to promote their tax policies, and perhaps to provide improved citizen tax experience. I am also optimistic that governments across the globe will leverage the most sophisticated AI tools to fight tax fraud—an annual multitrillion-dollar global problem.
Government authorities must also determine how these powerful AI solutions may prompt new laws and regulations. How should tax determinations recommended by AI tools, and adopted by human tax professionals, be audited? Will government officials attempt to regulate the algorithms that power tax decision-making? How will government authorities establish that no algorithmic biases exist that would unfairly burden a particular class of taxpayers? These questions and many others will be on the agenda of policymakers.
Lastly, I believe the development of AI tax technology tools should be shared across AI communities around the world. While there is significant AI development in the US and Canada today, other countries outside North America (and China in particular) will further invest in the lucrative AI race. This globally distributed AI development model will facilitate greater AI application to tax matters that vary by country.
The relevant question is not whether artificial intelligence technology will have an impact on taxes, but instead when will this powerful technology truly transform the field of taxation? While there are many questions around the promise of AI applied to tax, what I know for sure is that it is a very exciting time to be practicing within the tax profession.