What are your chances of being audited now that the IRS is using AI? Jury is still out
Current Audit Rates Remain Low
Recent data shows that fewer than 1% of taxpayers faced an audit in the past few tax years. For instance, in 2021, the IRS conducted audits on just 0.3% of all filings. While certain groups, such as those reporting specific income types or claiming particular tax benefits, historically have higher audit rates, these remain below 10% for most categories. Even within these groups, the proportion of individuals selected for review typically stays under 1%.
IRS Faces Major Personnel and Budget Shifts
Significant changes have occurred within the IRS over the past year. A substantial portion of its workforce has left, including many seasoned professionals in enforcement and complex audit operations. According to a July 2025 Treasury Inspector General report, nearly a quarter of tax examiners and revenue agents were among those departing. Meanwhile, a majority of the enforcement funding promised by the 2022 Inflation Reduction Act was retracted, and the Trump administration plans further reductions in IRS budgets next year.
AI Integration and New Goals
The IRS has been investing in modernization, including the deployment of artificial intelligence in key areas like enforcement. “Artificial intelligence (AI) and advanced analytics will help us pinpoint high-risk non-compliance cases with greater precision,” stated IRS CEO Frank Bisignano in written testimony before the Senate Finance Committee. His priorities include “strengthening the compliance agenda and enhancing collections beyond previous standards.”
Uncertainty Over AI’s Impact
It remains unclear how these changes will affect future audit probabilities. This uncertainty stems from factors such as whether AI will be applied wisely and whether enough trained staff will be available to handle audits flagged by the technology. “The success of AI depends on how it’s balanced with human oversight,” noted former IRS Commissioner Danny Werfel. “Without that, the system might generate errors or target the wrong returns.”
AI’s Potential to Streamline Audits
Barry Johnson, former chief data officer at the IRS, highlighted that AI could improve efficiency in identifying non-compliant filings. “AI provides a forensic edge in selecting the most relevant returns for review,” he explained. The technology can detect anomalies in tax returns, such as discrepancies in income reporting, which might signal fraud or simple mistakes. “It’s like equipping auditors with night-vision goggles,” Werfel added, noting AI helps isolate potential issues without overwhelming resources.
Challenges in Implementation
Despite AI’s advantages, its effectiveness hinges on having enough experienced staff to address the outcomes. If the IRS increases correspondence audits—focused on single-year issues like income underreporting—human reviewers will be critical to support filers who seek clarification. Similarly, more complex cases requiring field or office audits will demand trained personnel to navigate intricate details. With a significant loss of expertise in 2025, the agency may face a shortage of skilled staff to maintain audit quality. “AI can’t replace the knowledge lost,” Werfel emphasized, stressing the need for human expertise to ensure accurate assessments.
