IRA funding will solve tech shortfalls

Technological advancements. According to Kovacev, the new enforcement funds won’t only be used to bring on more employees. There will also be investments toward “sophisticated data analytics and artificial intelligence tools, which in the long run will probably be much more profitable than just adding a bunch of bodies.” He explained that advanced algorithms can be run by machine-learning tools that help IRS agents root out anomalies based on data patterns. This can help with discovering underreporting using data the IRS has on hand. An example would be a mismatch between aspects of a corporation’s tax return and their financial statement. “Certainly with this money, I would expect there to be an increase in that kind of technological advancement,” Kovacev said, adding that “a big chunk” of the appropriation will be deployed for current compliance enforcement in areas such as conservation easements, microcaptive insurance companies, and virtual currency. Recently published research presented at a June webinar co-hosted by the IRS and the Urban-Brookings Tax Policy Center suggests that artificial intelligence can better predict where to recoup owed taxes. A study released in April by a group of Stanford University and IRS researchers presented a possible opportunity for using machine learning to “optimize” audit selection. The IRS already uses a random selection model to choose which returns are audited based on noncompliance risk factors. The systems are automated and rely on filters to flag returns with an increased likeliness for error. Improved tools developed because of the reconciliation bill could perhaps refine this process. Beyond enforcement, the additional funds are expected to upgrade the IRS’ computer systems, further advance customer service software, and improve the overall taxpayer experience.