Tax evaders, BEWARE!
The Income Tax department is using AI data analytics to bust fake deductions, flag HRA fraud, and trace digital footprints.
Tax authorities in India have greatly expanded AI and data analytics to spot bogus deduction claims.
Officials now feed Statement of Financial Transactions (SFT) data from banks, mutual funds and registries into AI systems, cross-checking against ITRs, TDS, GST and other records.
As Anil Rego, founder of Right Horizons, a financial and wealth management firm explains, "They have access to a lot of data and are using AI and other tools to cross-check information from multiple sources. What's also changed is that earlier, different arms of the government didn't talk to each other. Now, there's much more integration across departments."
AI-driven trend analysis tools flag significant deviations between taxpayers' current and past filings, making audits more risk-based and efficient. This automated 'faceless' assessment reduces manual work and identifies suspicious cases quickly.
The tax department uses AI to match ITR deductions against third-party data. HRA claims via rent receipts are cross-verified using Form-16, AIS, and Form-26AS, with inconsistencies instantly flagged.
"When someone claims House Rent Allowance (HRA), and the rent paid is over Rs 1 lakh per annum, the PAN of the landlord must be submitted. If you're paying rent to a relative and claiming it, but that relative doesn't report the income in their return, it raises a red flag. The IT department can easily cross-check such mismatches," says Mihir Tanna, who works at a well-known chartered accountancy firm in Mumbai, and has more than 15 years of experience in direct taxation.
In 2023-2024, the department launched 'outreach' to 8,000 taxpayers whose donations exactly matched tax break thresholds. Similar AI checks are run on home loans, medical insurance and other deductions.
Every policy number, account number, or vehicle registration must be provided when claiming benefits, or the system spots suspicious entries.
As Anil Rego notes, "They are matching credit card usage and UPI transactions much more effectively now, creating digital trails they can follow."
AI/Data Tools & Third-Party Data
To tackle fraudulent tax evasion, the Income Tax department uses several data-driven systems.
The Annual Information Statement (AIS) compiles taxpayer transactions -- bank interest, mutual fund dividends, TDS, property purchases -- from multiple agencies.
The Statement of Financial Transactions (SFT) captures cash deposits, share trades and high-value property deals.
AI systems triangulate this data with filed returns.
For example, if a landlord's PAN shows no rent received via AIS but the tenant claimed large HRA, the mismatch triggers a notice.
Large payments flagged in GST or crypto are checked against business income.
This 'Project Insight' approach melding AIS, SFT, TDS/TCS creates a holistic analytics platform scoring risks on each ITR, says Rego.
The I-T department is integrating non-financial data too.
A recent initiative obtains passenger data from DigiYatra app and matches it against tax returns. High-frequency or high-cost travel not justified by reported income can expose evaders.
Officials note DigiYatra's traveler IDs are 'reconciled with income tax filings' to spot suspicious spending patterns.
While the media has widely reported and tax experts believe that the IT department accesses DigiYatra data for cross-referencing, the Government of India has strongly refuted these claims.
All ITR forms now require exact PANs of landlords or insurers and registration numbers, enabling AI to automatically verify claims.
As Anil Rego notes, "They are monitoring high-value credit card transactions. If you cross a certain threshold, that data is being tracked and cross-referenced."
Mihir Tanna adds, "The Income Tax department tracks IP addresses from which ITRs are filed. If many returns are filed from the same IP with similar deduction patterns, they use this to trace the physical location and uncover consultants running such operations."
Case Studies: Enforcement Actions
Several recent enforcement actions illustrate the approach and its success:
Salaried Employees (80G/80GGC): In early 2023, 8,000 taxpayers received notices for making exactly-threshold donations to charity/political trusts.
Data analytics revealed their gifts were exact amounts needed to drop into lower tax slabs. The department scrutinised trusts issuing receipts, threatening loss of tax-exempt status.
This targeted drive used Section 80G/80GGB/80GGC data cross-checked with donor incomes to spot fake receipt mills.
House Rent Allowance (HRA) Claims: The I-T department examines HRA claims via AI. In FY23-24, it flagged hundreds of rent receipts that didn't match Form-16/AIS data.
If employees claim over Rs 1 lakh rent (requiring landlord PAN) but AIS shows no payment to that PAN, alerts are generated. AI-based detection of fake rent receipts led to sharp rises in HRA-related notices in 2024-2025.
The recurring theme: AI identified unusual patterns among large employee groups, triggering targeted enforcement actions.
Expert and Official Insights
Tax officials emphasise a 'compliance-first' approach focused on data and outreach rather than harassment.
The CBDT encourages employees to amend errors before facing penalties, with campaigns run through employers to warn against fraudulent claims.
Common Fraud Patterns
Mihir Tanna identifies typical schemes: Claiming bogus deductions like fake housing loan interest or inflated LIC premiums. Under the old regime, people exploit fake personal loan interest deductions or pay rent to family members who don't report rental income.
Consultants facilitate fraud by targeting tax-unfamiliar individuals, charging Rs 2,000 to Rs 5,000 to falsely inflate deductions not reflected in Form 16. Recent raids have targeted these organisations helping people claim fake exemptions.
Advanced Scrutiny Methods
Anil Rego notes scrutiny has become highly targeted rather than random, using AI to flag suspicious transactions with specific triggers and criteria for more effective investigations.
The AI system spots inconsistencies by analysing spending patterns against past returns. For instance, property purchases are compared with five years of ITR data, including location, lifestyle-based expected savings, and reported income to identify discrepancies warranting scrutiny.
Dos and Don'ts for Taxpayers filing ITR for FY 2024-2025
Taxpayers must exercise extreme diligence when filing for FY 2024-2025 (AY 2025-2026) as scrutiny intensifies.
Essential Dos:
Reconcile All Data: Cross-verify AIS/Form-26AS against personal records before filing.
Ensure every large transaction -- bank interest, mutual fund units, property sales -- is accurately reported.
Check if deductions appear in 26AS or AIS (like donation receipts in trust's AIS), as mismatches trigger instant flags.
Maintain Comprehensive Documentation: Keep detailed paperwork for each claimed deduction.
Provide exact policy IDs for Section 80C investments, insurer details for medical insurance (80D), lender information with account numbers for loans (80E/80EE/80EEA), and vehicle registration for EVs (80EEB). These details are now mandatory -- missing fields cause queries or auto-rejections.
Verify HRA Claims: For HRA over Rs 1 lakh annually, secure landlord's PAN and ensure rent payments appear in AIS records. Avoid relying on handwritten receipts without matching bank trails, as ITD actively compares declared HRA with AIS data.
Report Complete Income: Declare full income honestly. Avoid splitting income across relatives or falsely reporting losses. High-value expenditure like foreign travel draw scrutiny if income hasn't grown correspondingly.
Use Official Channels: Respond to notices only through the e-filing portal's 'Responses' tab for genuine communications. Beware phishing calls promising easy refunds -- always verify through the portal.
Critical Don'ts
Never Claim Without Proof: Avoid deductions without valid documentation. Reject offers to file bogus deductions promising 100% refunds -- penalties now reach 200% for false claims.
Don't Ignore AIS Alerts: Address discrepancies in AIS or Form-26AS before filing. Sometimes banks report unknown TDS entries. As Mihir Tanna notes, "AIS is just a reference tool -- if actual income exceeds AIS figures, report the higher amount."
Don't Hide Foreign Assets: New ITR schedules mandate declaring overseas income/assets. Evasion attracts severe penalties.
Avoid Dubious Consultants: Steer clear of 'ITP' agents promising guaranteed refunds. Authorities investigate such intermediaries -- use reputable chartered accountants instead.
The I-T department's message is clear: Substantiate every claimed deduction. Voluntary compliance through official e-portals prevents penalties and audits. Transparency remains the optimal strategy.
© 2025