IRS pilots agentic AI tool to prepare fraud cases for staff

Nov 14, 2026
The agent would review taxpayer responses, summarise them and build a case before handing it to an employee for a decision, an IRS official said.

The IRS is piloting an agentic AI tool to handle the routine steps of potential fraud cases, Jim Clifford, director of return integrity and compliance services in the agency's Taxpayer Service Division, said at ACT-IAC's Service Delivery and Experience Summit, FedScoop reports.

The IRS already uses machine learning models to flag anomalies in tax returns. The agent would work after that point, organising correspondence with taxpayers and reading the documents they send. "What we want the agentic AI to do is review the response, summarize it, build a case, and hand it off to an employee at the point where it's really ready to make a decision," Clifford told FedScoop. He said "making judgments" will always stay with a person.

Clifford gave no timeline for deployment. He said the pilot is being developed in a separate workstream and will be scaled to other parts of the IRS "once it's tested and proven to be successful." He said the agency faces about 1 billion attempts each month by fraudsters to obtain unauthorised information. The Government Accountability Office estimates the federal government loses between $116 billion and $304 billion a year to tax fraud, according to FedScoop.

We hope you enjoyed this article

Consider subscribing to one of our newsletters like AI Policy Brief or Daily AI Brief.

Also, consider following us on social media:

AI Safety & Regulation AI Brief AI Brief (X)
Free newsletter

AI Policy Brief

Weekly report on AI regulations, safety standards, government policies, and compliance requirements worldwide.

Guide
10 most valuable public AI chip companies

10 most valuable public AI chip companies

Publicly traded companies designing merchant silicon for AI computing, ranked by market capitalization at the close of 23 October 2026.

Whitepaper

Tensordyne Napier: What If One Rack Could Do the Work of Nine?

Tensordyne

This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.

Read more

You may also like

Free, six days a week

Daily AI Brief: the AI news that matters, in your inbox.