New Report: IRS's Development & Use of Artificial Intelligence

Governance Efforts Should Be Accelerated To Ensure the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence

What is Artificial Intelligence, or AI?

AI are computing systems that “learn” how to improve their performance. AI has the potential to enhance the IRS's daily operations, improve efficiency, and provide more effective methods for tax administration. The IRS has been working with AI-like technology since 2008 and continues to expand its use.

Are there any risks to IRS using it?

The IRS must ensure that taxpayer rights are adequately protected. The use of AI has the potential to amplify existing biases and concerns related to civil liberties, ethics, and social disparities. Biases may arise when AI systems are created using data that may reflect preexisting biases or social inequities.

What did TIGTA find?

As of February 2024, the IRS reported 30 active AI projects and 38 planned AI projects aimed at enhancing operations, customer service, and enforcement.

Most of the projects are internal, and most involve contractors. 

• 18% of IRS AI projects are external projects that require taxpayer engagement. Most of these projects are either chatbots or voicebots that all taxpayers can access via the IRS.gov website or by calling and navigating an interactive voice response system.
  
  

• 68% of projects involve some level of contractor support. We found that base year contract costs exceeded $146 million with additional option years’ costs to be $621 million.

To strengthen its governance, the IRS has established a new AI oversight organization, led by the IRS Chief Data and Analytics Officer. The IRS noted that supplemental funding it received through the Inflation Reduction Act provides the IRS increased opportunities to expand its use of AI.

For more findings:

Having trouble viewing this email? View it as a Web page.