AI-Powered Tax Intelligence: CBDT’s Data Tools Help Uncover Hidden Wealth and Boost Tax Collection by Rs. 11,000 Crore

According to the Central Board of Direct Taxes (CBDT), Chairman Ravi Agarwal, for more than the previous four years, the use of various artificial intelligence (AI) tools by the Income Tax Department has facilitated over one crore taxpayers updating their returns, resulting in an additional Rs. 11000 crore in tax collection.

In an interview, the chairman emphasised the point of how the use of AI-powered data analytics tools has helped detect taxpayers who had either not filed their returns or had filed incorrect ones. He said, “The analysis involves two aspects, the AI tools and the data they analyse.” Distinct tools are located at various stages, including the Centralised Processing Centre in Bengaluru.

The government creates Annual Information Statements (AIS) for 40 crore individuals who have PAN cards; however, only about nine crore of them actually file income tax returns (ITRs). One of the key jobs of the AI system is to check if people who did not file tax returns should have filed them, based on their past data and financial transactions.

The system also reviews the tax returns that people have already filed to check if they are correct and to catch repeat defaulters. This analysis is part of the department’s NUDGE framework (Non-Intrusive Usage of Data to Guide and Enable), where they send advisory messages or letters to taxpayers. These messages ask people to either correct their tax returns or confirm that the details they filed are right. Because of this approach, over one crore updated tax returns have been filed since 2021-22.

This strategy adopted by the department has led to the extension of time to update ITRs to four years from two years. The chairman said, “If a discrepancy is being flagged, the taxpayer should have a window to make corrections within the statute.”

AI-based actions have also helped find hidden wealth kept outside India. In the financial year 2024-25, the department found foreign assets (property and money) worth Rs. 29,000 crore that were not reported earlier. They also found more than Rs. 1,000 crore in foreign income linked to cryptocurrencies and other virtual digital assets (VDAs).

According to a recent campaign held between January and March, under Section 80GGC, the department used a similar AI-based mechanism to check claims of political donations. Around 6.25 of taxpayers whose claims do not match the data from political party contribution reports filed to the Election Commission were sent alerts via SMS and email. An additional Rs. 404.2 crore has been added to tax collection, in conclusion to around 35,260 taxpayers changing their returns.