₹10,000 Crore, 95 Lakh "Trainees," and a Paper Trail That Falls Apart on Contact

The government's flagship skilling scheme has a data problem, and the numbers are not small. Pradhan Mantri Kaushal Vikas Yojana was launched in July 2015 to give India's youth industry-relevant skill certification. Between 2015 and 2022, the scheme released ₹10,194 crore and certified 1.10 crore candidates across three phases. On December 18, 2025, the Comptroller and Auditor General tabled its performance audit — CAG Report No. 20 of 2025 — in the Lok Sabha. What it found does not just flag inefficiency. It raises the question of whether large parts of this data were ever real. The bank accounts that weren't Of 95,90,801 participants certified under PMKVY 2.0 and 3.0, the scheme's own records show that 90,66,264 — 94.53 percent — had bank account fields that were blank, zero, "N/A," or null. That's out of a scheme built to pay a ₹500 Direct Benefit Transfer to every certified candidate. Of the roughly 5.25 lakh candidates who did have something entered, 12,122 unique account numbers were reused across 52,381 people — the same number appearing for two candidates in some cases, and for more than 2,000 different people in one instance. Where individual accounts were listed, the audit found entries like "11111111111," "123456," single-digit numbers, or plain text where a bank account number should be. Contact details tell the same story, at a larger scale The audit also examined email and mobile fields for the same candidate pool. Here the picture is worse than headline coverage has generally captured: Over 71 lakh candidates (71,20,995) shared an email address with at least one other candidate — a single email address was linked to more than 5,000 different people in 53 cases. Another 36.49 lakh candidates simply had "Migrated Data" recorded as their email. On the mobile side, over 10.5 lakh candidates (10,56,475) shared a phone number with someone else, and roughly 87,000 numbers were invalid outright — fewer than 10 digits, or strings like "1111111111." As of October 2024, only 63.75 percent of certified candidates had actually received their DBT payment. More than 34 lakh certified candidates are still owed money. Beyond the data fields The report also documents irregularities that go past bad data entry. Sector Skill Councils tasked with certifying training in some cases doubled as Training Providers receiving the funding — a structure the audit flags as a conflict in the RPL-BICE ("Best-in-Class Employer") component, where oversight was weakest. Multiple news outlets, citing the same report, describe photographic evidence — batch photos required for certification — being reused across different states and different candidates, and training agencies claiming implausibly large numbers of certified trainers relative to the actual job openings they were tied to. Physical inspections had their own credibility gap: one CAG-documented case shows a single inspector recording physical site visits across eight states — Andhra Pradesh, Gujarat, Jharkhand, Karnataka, Madhya Pradesh, Maharashtra, Tamil Nadu, and Telangana — in a single day. The larger picture Overall placement under the scheme's core Short-Term Training component stood at just 41 percent. The audit also found no long-term skilling strategy was ever prepared, no alignment between what was trained and what the labour market actually needed, and weak convergence between PMKVY and dozens of parallel skilling schemes run by other ministries and states. The Ministry's response, on record, points to PMKVY 4.0 reforms — Aadhaar-linked e-KYC, geo-tagged attendance, and a new Skill India Digital Hub — as fixes going forward. Whether those close the gap the CAG has now documented in writing is the next thing worth watching. For the full audit trail and more from NavvyaSignal's India Desk, visit www.navvyasignal.com