The Centre has rolled out a tougher, AI-led verification layer for attendance under its flagship rural employment programme, the Viksit Bharat–Guarantee for Rozgar and Ajeevika Mission (VB-G-RAM-G). Union Rural Development Minister Shivraj Singh Chouhan announced on 23 October 2026 that an upgraded version of the National Mobile Monitoring System (NMMS) app, used to mark workers present at worksites, will now carry liveness detection, anti-tampering safeguards and refined face verification.
According to a statement released through the Press Information Bureau, the minister said the government’s priority is to protect workers’ wages while making sure public money is spent transparently and for its intended purpose.
Why the upgrade was needed
Face authentication was first introduced in March 2026, in response to concerns over fake attendance under MGNREGA, the scheme that preceded VB-G-RAM-G. Field checks, however, continued to surface gaps: photographs of photographs being uploaded instead of live images, mismatches between actual and recorded headcounts, inconsistencies in the gender profile of workers at some sites, the same picture appearing across multiple muster rolls, and differences between morning and afternoon photo sets.
What changes on the ground
- Liveness detection and anti-cloning: the app is designed to confirm that the person in front of the camera is a real worker present at the site, not a printed image or an old photograph.
- Region-aware face matching: AI-based face verification has been tuned to account for variations across states and geographies.
- AI photo de-duplication: a module flags suspicious or repeated images. Where such entries are detected, the wage list and the Fund Transfer Order (FTO) are automatically held back until checks are complete.
- Offline attendance: worksites without network connectivity can still record attendance offline.
- Authorised exceptions: a fallback mechanism covers elderly workers, persons with disabilities, or device failures where face authentication does not work.
Scale of the programme
Since its launch on 1 July, VB-G-RAM-G has 31.23 lakh active worksites and has provided employment to more than 3.65 crore workers from 2.89 crore rural households, the minister said. Close to 46 crore person-days have been generated so far, of which about 40 crore were recorded through face authentication.
The governance angle
The upgrade is a clear example of how digital public infrastructure is being used to tighten leakages in direct benefit programmes. By linking verified attendance to payment release, the ministry is moving checks upstream, before money leaves the system, rather than relying on audits after the fact. The inclusion of offline capture and an exception route also reflects a recognition that automation in rural welfare must stay workable for the last mile, where connectivity and device quality vary widely.
For states and district administrations, the next test will be execution: how quickly flagged cases are resolved so that genuine workers are not left waiting for wages, and how transparently exception approvals are monitored.
