AI and Automation in Tax

This post first appeared on IBM Business of Government. Read the original article.

Monday, January 30, 2023

Blog Co-Author: Peggy Bogadi, Former Commissioner, Wage and Investment Division, IRS

In late 2022, we kicked off a series of blogs focused on the role of technology in modernizing tax agencies around the world. Leaders recognize that technology is critical to improving taxpayer service, closing tax gaps, and addressing workforce challenges. Those executives share many ideas with their peers across governments and the commercial world.

Morning Consult recently surveyed 4,000 global business leaders about their 2023 investment plans. Despite forecasts for a leaner year and potential global recession, the survey found that organizations are continuing to invest in technology. Technology is essential for cost-effective workload growth and scale for governments and other organizations. It is a crucial source of competitive advantage for companies.

High Impact Technologies

Those responding to the survey cited digital employees (35%) and generative AI (35%) as the top emerging technologies that will bring fresh innovation to their organizations over the next 3 to 5 years.

Digital employees will augment how workers manage their day-to-day work. AI-powered assistants will increase employee productivity. They will support activities in human resources like job changes and department transfers. They will automate finance and invoice processing, and speed up IT support. As Generative IT moves from interesting experiments like those of ChatGPT to enterprise technology offerings, more organizations will be able to deploy AI in a wider range of mission-critical situations.

AI and Automation is critical for Tax agencies. Artificial intelligence is already improving taxpayer service. Countries and regions around the world increasingly use virtual assistants to answer taxpayer questions — reducing the burden on call centers and improving service. As discussed in our recent blog on taxpayer engagement, Spain has been using Virtual Assistants to answer questions about Tax Law Changes Affecting both Corporations and Individuals. After a 2018 change to the Value Added Tax process, a virtual assistant handled thousands of questions. It reduced e-mail traffic between corporations and the tax agency by about 80%.

AI and Automation at the US Internal Revenue Service

In the US, the Internal Revenue Service (IRS) is focused on improving taxpayer experience and reducing fraud and error to close the tax gap. The IRS will use both people and technology to do it. Both are critical to an agency where the average worker is 55 years old and about ½ of employees will reach the end of their careers in the next few years. Further, IRS:

  • Full-Time Equivalents (FTEs) fell by 1/3 from 119,000 in 1991 to 79,000 in 2021
  • Audits dropped 58% between 2010 and 2020 and the IRS has the same number of auditors today that it did in 1954
  • Offices are increasingly engaged in new missions like health and social programs
  • Collections increased by nearly 4x from $1.1T in 1991 to $4.1T in 2021

AI and Automation have an important role in improving taxpayer service and closing the tax gap – the difference between taxes owed and taxes paid. The IRS has begun using these technologies and demonstrating that they can deliver value. In the first ½ of 2022, the IRS began using voicebots that allow callers to navigate the IRS phone system to receive a faster response. IRS voicebots answered over 3 million calls and it is adding new functions for taxpayers to resolve issues. For example, taxpayers can use the bots to set up most types of payment plans.

In addition to bots when discussing AI and automation, in late 2022 the IRS used intelligent automation to digitize thousands of paper tax returns. While 92% of individual returns are e-filed, nearly 13 million taxpayers filed paper returns. Refunds for e-filers generally take a few weeks to process, whereas in 2022 refunds for those filing paper returns could take as much as six months. During the pandemic, there were huge backlogs of paper tax returns. Taxpayers filing paper returns waited months for someone at an IRS Service Center to open and process the tax return.

The IRS demonstrated that intelligent automation could speed paper processing. In October 2022, the IRS used new processes and technology to scan paper returns, capture the data, and post it directly into IRS systems. Only the returns with errors were touched by employees. Paper returns were processed in days not months – allowing refunds to be sent to taxpayers nearly as quickly as taxpayers who file electronically.

Automating Paper-Based Processes at US Veterans

Beyond processing paper tax returns, the IRS and other tax agencies could use AI and automation to streamline processes for tax adjustments, hiring support, and improving employee engagement. For example, in December 2019, the US Department of Veterans Affairs awarded a contract to streamline incoming correspondence — including paper mail, faxes, email, and benefits applications. In 2019, 1000 Department staff were involved in moving paper across hundreds of programs to serve veterans. On average, mail took 27 days to be opened, read, and moved to the right place for disposition. The Department wanted to increase efficiency and better serve veterans.

Within 45 days of contract award, the Department had a team and technology in place to process the first transactions. After just 18 months, 950 of the 1000 Department staff involved in managing incoming correspondence were trained for new, higher-value roles. More than 280 types of correspondence had been automated and cycle time fell from 27 days to 24 hours.

What’s Next?

The opportunities for AI and Automation in Tax are real. In a future blog, we will dive more deeply into the IRS intelligent automation proof of concept. We will look at the approach, benefits, and applicability to other Tax agencies around the world.

Image by xb100 on Freepik

Leave a Reply

Your email address will not be published. Required fields are marked *