IRS Explores AI to Tackle Identity Theft Case Backlog

IRS Considers AI for Identity Theft Backlog Amidst Long Wait Times

The Internal Revenue Service (IRS) is exploring the use of artificial intelligence (AI) to manage a significant backlog of identity theft cases. As of July 25, 2026, the agency faced nearly 257,000 unresolved cases. This backlog has led to taxpayers waiting an average of 533 days for their cases to be assigned to an employee, with overall resolution taking close to 20 months. This is far beyond the IRS’s own goal of resolving such cases within 120 days. The agency is currently assessing whether AI can help sort and route these complex cases more efficiently, but no AI system has yet been put into action for this specific problem.

The Longest Delays Occur Before Case Assignment

A major issue identified by the Treasury Inspector General for Tax Administration (TIGTA) is the extended period cases spend waiting before an IRS employee begins working on them. The average wait time for a case to be assigned is 533 days, meaning taxpayers often wait over a year and a half before any substantive work is done on their claim. While the total resolution time is often cited as around 20 months, this includes the time spent by an employee actively working on the case. Once assigned, some cases are resolved more quickly, with the agency completing a majority of taxpayer-reported cases within 30 days of assignment. However, these faster resolutions do not compensate for the lengthy initial waiting period.

Backlog Size Shrinks but Remains Substantial

The number of open identity theft cases has decreased over the past few fiscal years, according to TIGTA reports. For example, the backlog stood at 485,313 cases at the end of fiscal year 2023, dropping to 315,740 by the end of fiscal year 2025. The most recent count of 256,988 cases as of July 25, 2026, shows a continued reduction. This backlog grew significantly during the pandemic due to an increase in identity-related tax fraud. Despite the overall decline in case numbers, the average resolution time has remained stubbornly high, hovering around 20 months. This indicates that even with fewer cases, the system is still struggling to process them within the agency’s target timeframe.

Watchdog Report Highlights Missed Targets

A review conducted by TIGTA of 114 closed identity theft cases revealed that, on average, these cases took 655 days to resolve. Out of this sample, only four cases were completed within the IRS’s 120-day target. This finding underscores the significant gap between the agency’s stated goals and the actual time it takes to resolve these complex issues. The extended timelines not only cause frustration for taxpayers but also result in financial costs for the government. TIGTA calculated that the IRS paid approximately $124.2 million in refund interest due to these delayed resolutions between fiscal years 2023 and 2025.

Management Considering AI for Triage

IRS management, including Kenneth Corbin, chief of the Taxpayer Services Division, has acknowledged the need for faster case processing. However, they have expressed reservations about adding another manual screening step before assignment, believing it might not improve timeliness with current staffing levels and could lead to duplicated work. Instead, the agency is looking at AI as a potential solution to assess case complexity and assist in routing them to the appropriate staff. This approach aims to streamline the initial stages of case handling without requiring additional manual review by already busy employees. The agency’s next steps will depend on the outcome of its assessment into AI’s capabilities for this task.

Financial Impact of Delayed Refunds

The substantial interest paid on delayed refunds represents a direct financial consequence of the lengthy identity theft case resolution times. The $124.2 million figure calculated by TIGTA for fiscal years 2023 through 2025 is directly linked to cases that can spend hundreds of days in the queue before being assigned. This highlights the economic as well as the taxpayer service implications of the current backlog. The IRS’s decision on whether to implement AI for case triage will be informed by its evaluation of how effectively the technology can improve the sorting and routing process, ultimately aiming to reduce both taxpayer wait times and associated government costs.

Frequently Asked Questions