The size of the problem
Industry studies place credential discrepancy rates in Indian CV samples between 0.5% and 2% across hiring volumes. Translated to a national hiring scale of millions of offers, that's tens of thousands of fabricated credentials reaching enterprise pipelines every year. The cost compounds: a fake-degree hire eventually exits, the role re-opens, the BGV process starts again — and the candidate's wider behaviour profile (other CV claims, other employers) gets re-validated by the next hiring team.
The good news: most fakes are not sophisticated. They fail one of twelve forensic checks, each of which a trained HR reviewer can perform in under a minute. The harder problem is workflow scale — running these checks consistently across hundreds or thousands of offers per cycle.
The 12 forensic signals
Walk every candidate's academic claims through these twelve checks. Most fakes fail two or three of them simultaneously.
1. Mark-sheet aggregate doesn't reconcile
Sum of subject marks ÷ count of subjects ≠ stated percentage. The single most common error in fabricated mark-sheets. Forgers update the headline percentage without re-doing the arithmetic on subject scores.
2. University issued before the institution opened
Cross-check the issue year against the university's founding year and accreditation date. Both are public on UGC's recognised-universities list and the institution's own website. A 2019 degree from a university gazetted in 2021 is a definite fail.
3. Programme structure violates UGC norms
A "B.Tech in 3 years" or "M.Tech in 1 year" from a non-deemed Indian university is almost certainly fabricated. UGC programme durations are standardised (B.Tech: 4 years; M.Tech: 2 years; B.Sc.: 3 years). Foreign-tied programmes have specific accreditation pathways — verify those separately.
4. Font and seal inconsistencies
Genuine certificates use offset printing on watermarked paper with embossed seals. Inkjet output, slightly mis-aligned text, or pixellated logos when zoomed are red flags. Also: the registrar's signature on a single candidate's mark-sheet should match the signature on other mark-sheets from the same university and year.
5. Roll number outside published range
Most universities publish their result archives. If the roll number falls outside the cohort's published range, the credential needs registrar confirmation. A roll number that doesn't exist in any published cohort is a hard fail.
6. Issuer mismatched to programme
A research-track PhD from an institution without research accreditation is implausible. UGC and NAAC databases are public — check the institution's grade and approved programme list before accepting unusual claims.
7. No NAD record (post-2020)
Most UGC-recognised programmes push results to NAD post-2020. A credential claiming to be from a participating institution but absent from NAD warrants a registrar letter. Note: NAD coverage is not universal yet, so absence is a flag, not a definite fail.
8. DigiLocker pull fails for a participating issuer
Ask the candidate to fetch their degree via DigiLocker live, in front of the interviewer. If the issuer is a DigiLocker participant and the pull fails, the credential is suspect. Easy, fast, candidate-cooperative check.
9. Email domain on signed PDF doesn't match issuer
A signed PDF whose certificate authority chain points to an unrelated domain (free-tier signing services, or entirely different organisations) suggests the document was self-signed by the candidate, not the registrar.
10. Date inconsistencies across CV claims
Candidate claims a degree completed in 2022 but lists full-time employment that overlaps with their final semester in 2021–22. Either the degree dates are wrong or the employment dates are. Both can be falsified; cross-check both with the source.
11. Course list inconsistent with programme
A B.Tech transcript listing courses that don't belong to that programme's syllabus, or missing core courses that every student in the cohort would have taken. UGC syllabi are standardised; AICTE programmes have published model curricula. Compare.
12. The candidate cannot answer programme-level questions
This is the cheapest, fastest check. If a candidate claims a degree in a specific subject from a specific institution, they should be able to name a faculty member, a core course, or a campus location. A 30-second conversation catches a lot of fake credentials before any document forensics begin.
The four legitimate verification routes
For candidates who pass the forensic checks but still need definitive verification, four routes exist:
- Manual UGC / university verification. 2–8 weeks, ₹500–₹2,500 per request. Required when an embassy or foreign university demands a registrar-signed letter.
- DigiLocker pull. Instant, free, but only works for participating issuers. Best first check for post-2020 graduates.
- BGV vendor (AuthBridge, IDfy, OnGrid, etc.). 3–14 days, ₹150–₹800 per check. Standard for corporate hiring at 100+ offers per cycle.
- Public blockchain verifier. Under 10 seconds, free, unlimited. Works for any credential issued on a participating credential network — coverage is growing as universities onboard.
The full comparison — including coverage gaps, common failures, and when to use which — is on the /verify-indian-degree-online page.
Designing the verification workflow
At hiring volume, the question stops being "how do I verify one degree" and becomes "how do I verify 500 degrees per month consistently, with audit trail". The answer is a tiered workflow:
Tier 1 — Automated checks at offer stage
ATS-integrated forensic checks: aggregate reconciliation, programme duration validation, university existence check, DigiLocker pull. These run automatically when the candidate uploads documents — no human time. Roughly 70% of pipeline reaches Tier 2 cleanly; the rest get flagged for review.
Tier 2 — Human review of flagged files
A trained HR reviewer takes 5–10 minutes per flagged file to walk through the remaining forensic signals and decide whether to escalate. Most flags resolve cleanly — a percentage that reconciles after re-OCR, a roll number that turns out to be in the published range. Genuine edge cases proceed to Tier 3.
Tier 3 — External verification
For the small percentage that still need definitive verification: BGV vendor for standard cases, registrar letter for sensitive roles, blockchain verifier where applicable (which makes 60–80% of Tier 3 effectively free).
Why blockchain credentials change the math
The forensic checks above are necessary because most Indian degrees are PDFs. PDFs cannot prove their authenticity to the verifier independently — the verifier has to trust either the issuer's certificate, the registrar's phone confirmation, or a BGV vendor's database lookup.
A blockchain-anchored W3C Verifiable Credential is different. The proof of authenticity travels with the credential, mathematically. A 10-second public-verifier check confirms the issuer, the holder, and the integrity of every field — with no phone calls, no vendor fees, no ambiguity.
For credentials issued on a participating network, forensic signals 1–11 above become unnecessary — the cryptographic proof either succeeds or fails. The verification surface collapses from a 12-point checklist to a single API call. As more Indian universities adopt this infrastructure, the operational cost of HR verification at scale falls dramatically.
For BGV firms specifically
If you operate a BGV business, the smart play is to layer blockchain-credential verification on top of your existing manual flow. For credentials issued on a participating network, your verification cost goes to zero and turnaround drops to seconds — your margin per file improves, you compete on speed, and you free up your manual operations capacity for the cases that genuinely need it.
Treat blockchain verification not as competition, but as the layer that makes your BGV service faster and cheaper to operate. The verifications you keep doing manually are the high-value ones — sensitive roles, foreign credentials, edge cases — exactly where your operations team adds the most value.
Closing
Fake degree detection is a workflow problem, not a forensic mystery. Most fakes fail two or three of twelve simple checks. Build the workflow tier-by-tier, automate the easy tier, train the humans for the middle tier, and escalate intelligently for the rest. As more Indian credentials move to blockchain-anchored Verifiable Credentials, more of the pipeline collapses into the automated tier — and HR teams stop spending hundreds of person-hours per cycle on what is fundamentally a cryptographically solvable problem.