Fake ITR Detection for Self-Employed Loans: Verify Returns
Last reviewed: November 2026
Fake ITR detection for self-employed loans is the process of confirming that a borrower's income tax return (ITR) was actually filed, has not been altered, and reflects real business income before a loan is sanctioned. Lenders do it with PDF forensics, filing validation on the income tax portal, and cross-checks against tax, GST and bank data.
This guide is for CROs and heads of credit for MSME and self-employed lending at NBFCs, banks and fintech lenders. It covers how tax returns are faked or inflated, the checks that expose each pattern, and how to build them into underwriting without slowing good borrowers.
Key Takeaways
- For self-employed borrowers, the ITR is often the main income document, so an edited return distorts eligibility, limit and pricing at once.
- The e-filing portal's pre-login ITR status service works with an acknowledgement number and an OTP to any valid mobile number.
- The Annual Information Statement (AIS) shows GST turnover under code EXC-GSTR3B, while Form 26AS now covers only tax deducted or collected at source.
- A genuinely filed return can still be inflated, so filing checks must sit alongside GST and bank-credit checks.
Why Do Fake ITRs Distort Self-Employed Underwriting?
Fake ITRs distort self-employed underwriting because the return is usually the only document that states annual income, and every ratio downstream depends on it. Unlike a salaried borrower, a trader or professional has no payslip or employer to cross-check.
That makes the ITR the anchor for the fixed obligations to income ratio (FOIR), the share of income already committed to EMIs. Lift declared profit and the same borrower qualifies for a larger loan against property (LAP) or business loan.
The risk compounds in secured lending. A comfortable loan-to-value ratio protects principal, but an overstated income still produces a borrower who cannot pay, and recovery through the collateral takes years.
How Are ITRs Faked or Inflated?
ITRs are faked or inflated in six ways: editing a genuine ITR PDF, fabricating a return that was never filed, filing or revising a return just before applying, inflating turnover beyond GST filings, forging CA-certified financials, and parking one-off income in a single year. Only the first two are forgeries in the strict sense.
| Pattern | What Gets Changed | Where It Usually Shows |
|---|---|---|
| Edited ITR-V or computation PDF | Income, turnover, tax paid | Metadata, fonts, figures that differ from the portal copy |
| Return never filed | The whole document | Acknowledgement number fails on the portal |
| Filed or revised just before applying | Nothing forged; income is inflated | Filing date, revised status, lump-sum tax paid before filing |
| Turnover above GST filings | Turnover and profit | Gap between ITR turnover and GST turnover in AIS |
| Forged CA financials | Audited profit and loss, balance sheet | Missing or non-matching UDIN |
| One-off income in one year | Nothing forged; income is lumpy | Sharp jump against earlier years and bank credits |
The last three patterns pass file forensics. That is why ITR checks must reach outside the document.
How Do Lenders Detect Fake ITRs Before Sanction?
Lenders detect fake ITRs with four layers of checks, run from the file outward. Each layer catches a pattern the others miss.
- File forensics: confirm the ITR-V, computation and financials are unedited PDFs.
- Filing validation: confirm the return exists on the portal with the same figures.
- Tax footprint: confirm AIS and Form 26AS support the declared income.
- Business reality: confirm GST returns and bank credits match the turnover.
Layer 1: Has the ITR File Been Edited?
The ITR-V is the acknowledgement generated after a return is filed. Edited copies often carry a consumer editor in the producer field, a modification date after the filing date, or figures set in a different font from the surrounding text.
Run the same checks on the computation of income and on any CA-prepared financial statements. Hash every file too, because the same template appearing across unrelated borrowers points to an organised source.
Layer 2: Was the Return Actually Filed?
Every filed return has an acknowledgement number printed on the ITR-V. The e-filing portal's pre-login ITR status service accepts that number with a valid mobile number for an OTP, so a fabricated acknowledgement fails quickly.
The stronger check is the source copy. With the borrower's consent, have them log in during an assisted session and download the filed return, then compare it figure by figure with the uploaded PDF.
Check the Unique Document Identification Number (UDIN) on any certificate or audit report from a chartered accountant (CA). ICAI's UDIN portal lets banks and other third parties confirm that a practising CA actually issued it.
Layer 3: Does the Tax Footprint Support the Income?
The AIS shows information reported by third parties, including tax deducted at source (TDS), specified financial transactions and, per the income tax department's FAQ, GST turnover from GSTR-3B. Form 26AS, the annual tax credit statement, has covered only TDS and TCS from AY 2023-24.
A trader declaring ₹2.4 crore turnover whose AIS shows half that in GST turnover needs an explanation. So does a professional whose TDS-reported receipts are a fraction of declared fees.
Layer 4: Does the Business Look Like the Return?
GST returns and bank credits show whether the business is as large as the return claims. Pull GST data as described in GST-based MSME lending, and fetch current accounts through an account aggregator for lenders.
Compare at least three years of returns. Steady growth is normal, but a profit jump that coincides with the loan application deserves a question, and Digiqt's income verification AI agent shows how returns, bank data and GST can be reconciled automatically.
How many self-employed files did your team approve last quarter on an ITR that was never checked against the portal?
Talk to Our Lending Technology Specialists
Visit Digiqt to see how DocuForge checks uploaded tax return PDFs for tampering before they reach underwriting.
How Does Fake ITR Detection for Self-Employed Loans Work on a Real File?
On a real file, the layers combine into one verdict on the income figure. In this illustrative scenario, a wholesale trader applies for a ₹45 lakh LAP over 15 years at 11%.
The uploaded FY 2025-26 ITR shows turnover of ₹2.4 crore and business income of ₹24 lakh, or ₹2,00,000 a month. The new EMI is about ₹51,100, existing EMIs are ₹30,000, and FOIR is about 41%.
| Check | Finding |
|---|---|
| Metadata | ITR-V and computation were modified in a consumer PDF editor after the filing date |
| Portal copy | The return downloaded in an assisted session shows turnover of ₹1.15 crore and income of ₹11.5 lakh |
| AIS | GST turnover is about ₹1.15 crore, matching the portal copy |
| Bank credits | 12-month business credits of about ₹1.2 crore through AA |
On the true income of about ₹95,800 a month, FOIR is about 85%. The file should be declined, and the sourcing direct selling agent's (DSA) other recent files reviewed.
The lesson for leadership: the edited figures were consistent with each other and looked professional. Only the portal copy, AIS and bank data told the truth.
Which Verification Approach Fits Your Lending Model?
The right approach depends on ticket size, collateral and the borrower's GST status. Larger and unsecured exposures need every layer, while small secured loans can use a lighter set.
| Situation | Recommended Approach |
|---|---|
| Small-ticket business loans to GST-registered firms | GST and AA bank data first, ITR forensics as a supporting check |
| Unsecured loans to professionals | Portal copy of the return, AIS receipts and bank credits |
| LAP for traders and manufacturers | All four layers, plus loan against property underwriting cash-flow checks |
| Non-GST micro businesses | Bank credits and multi-year ITR trend, with field verification |
What About Borrowers Who Declare Low Income Deliberately?
Many small businesses under-declare income for tax reasons, then ask lenders to accept higher "assessed" income. That is a credit policy decision, not a fraud check, and it should be made openly through surrogate programmes.
The danger is mixing the two. If assessed income is allowed, record it as such, and never let a borrower solve the gap by editing the return.
How Should ITR Checks Fit Into the Credit Workflow?
ITR checks should run at upload, with filing validation completed before the file reaches a credit manager. Three design choices matter.
First, score every uploaded return automatically. Digiqt's DocuForge runs 12 layers of forensic checks on uploaded PDFs, including metadata verification, incremental update tracking, trusted PDF source validation and hash integrity, and combines them into a document risk score.
Second, make the portal copy or AIS mandatory above a ticket threshold your board sets. Third, feed the score and any income gap into your credit decision engine so high-risk files cannot reach sanction.
Watch the channel as well. Digiqt's lending fraud detection AI agent shows how repeated templates and devices can be linked across applications.
Which Metrics Show ITR Checks Are Working?
Five metrics show whether ITR checks are working: portal validation rate, ITR-to-GST gap rate, flag rate by channel, confirmed fraud rate and early default on self-employed files. Track them by product and DSA.
| Metric | Formula | What a Weak Number Signals |
|---|---|---|
| Portal validation rate | Returns validated on the portal ÷ returns used for income | Teams are relying on uploaded PDFs |
| ITR-to-GST gap rate | Files with ITR turnover over 25% above GST turnover ÷ GST-registered files | Inflated returns are entering the book |
| Flag rate by channel | Flagged ITR files ÷ files received, by DSA | One partner far above peers |
| Confirmed fraud rate | Confirmed cases ÷ flagged files | Rules too noisy, or reviewers clearing too easily |
| Early default on self-employed files | First-six-EMI defaults ÷ disbursed files | Income is being overstated |
Which Mistakes Let Fake ITRs Through?
Four mistakes let most fake ITRs through, and all of them are fixable in policy.
- Treating the ITR-V as proof of filing. It is a PDF like any other until the acknowledgement number is validated.
- Reading one year in isolation. Inflation shows up as a break in the trend.
- Ignoring GST. For registered businesses, GST turnover is the most current independent check on the return.
- Accepting files from intermediaries. Returns forwarded by a DSA or consultant have no chain of custody.
What Is Changing Next?
Tax data is getting richer and easier to reconcile. AIS now carries a wide set of third-party information, including GST turnover, and the Income-tax Act, 2025 brings renamed forms from tax year 2026-27.
Update extraction templates and checklists for the new names, and build portal validation into your digital lending platform. For more on segment-specific automation, see AI agents for SME lending.
Want every self-employed borrower's tax return scored and validated before credit opens the file?
Talk to Our Lending Technology Specialists
Visit Digiqt to see how DocuForge and our lending platform stop income document fraud before sanction.
Sources
- Income Tax Department: FAQs on Annual Information Statement (AIS)
- Income Tax Department: Know Your ITR Status User Manual
- RBI: Non-Banking Financial Companies – Fraud Risk Management Directions, 2026
- Business Today: Form 16 to Form 130, New Income Tax Act Brings in New Forms From April 1 (25 March 2026)
