Summary

  • Only 33.20% of BharatNet’s service-ready Gram Panchayats were operational in November 2025; service-ready does not mean a household is connected.
  • Fibre faults, a 17-day average restoration time and unused capacity make repair ownership and service metrics the next accountability test.

A network can pass a construction milestone and still fail the person who needs a connection. BharatNet’s latest Comptroller and Auditor General (CAG) audit makes that distinction unusually clear: by November 2025, 2.19 lakh Gram Panchayats (GPs) were marked service-ready, or 96.70% of the revised 2.26 lakh scope. Yet only 33.20% of those service-ready GPs were operational.

Those figures describe different states. “Service-ready” is a project status; “operational” is a narrower test of whether the network is working. Neither number says that every household has a subscription, that every school is connected, or that a resident has measured a particular speed. The audit’s point is more useful than a single coverage headline: the asset can be present while its service is unavailable.

The cost of a repair queue

The CAG attributes about 48% of non-operational cases to fibre faults or cuts and reports an average 17 days to restore a fibre fault. That is not a minor maintenance detail. A long restoration interval changes the value of a backbone for every downstream provider that must sell a dependable service over it. The audit also says major telecom service providers withdrew over service-quality concerns.

The economics follow a chain. A funded build puts fibre in place. A fault interrupts traffic. A slow repair extends the interruption. Providers then have less reason to rely on the asset, while public institutions and residents wait for a service that the construction ledger has already counted as ready. An asset can be physically complete and commercially difficult to use at the same time.

The audit reports another measure that should remain separate: only 17.91% of available bandwidth was used. It also found dark fibre left unleased in some cases despite demand, citing missing policy or delayed decisions at State Implementing Agency level. That is not the same as a fibre cut. It points to a second operating boundary: even where capacity exists, commercial access may depend on a local decision and a workable offer.

The denominator matters. The 17.91% refers to available bandwidth, not the share of villages connected or residents online. The 33.20% refers to operational GPs among those service-ready. The 48% refers to non-operational cases attributed to fibre faults or cuts. Mixing the three would turn a useful diagnosis into a misleading claim about rural households.

A baseline, not a verdict on the new model

The audit covers BharatNet Phase II and uses a November 2025 status point. The current BharatNet portal describes the programme as a ring-topology fibre network with service-level-assurance-based operations and maintenance, and now presents FTTH, bandwidth and dark-fibre services through a demand process. The Digital Bharat Nidhi (DBN) FAQ identifies BharatNet and the Amended BharatNet Programme as ongoing projects.

That newer operating language matters, but it does not erase the audit’s baseline. Nor does the audit prove that the current assurance model has failed: the available evidence does not compare post-reform performance with the earlier status point. The relevant question is whether the newer model can publish operational results that move beyond construction readiness: availability, fault age, restoration time, repeat faults, provider participation and end-user connections.

BharatNet has reached considerable physical scale. The CAG recorded ₹39,888 crore spent against an approved ₹42,068 crore and described the infrastructure achievement as substantial. That matters. The strategic choice is not whether to value the build; it is whether the next public rupee and the next management review treat reliable operation as an equally visible output.

A service portal that accepts a location-based demand can help route requests, but it is not proof of fulfilment. A request may reveal demand; a confirmed installation, working connection and continuing uptime reveal delivery. The system needs to report those stages separately. Otherwise, the public record can show a ready GP without showing whether a provider can lease capacity, a technician can reach a fault, or a household is actually served.

Sources

  1. CAG, Performance Audit on BharatNet, Report No. 19 of 2026
  2. CAG report mirror, SAI India
  3. Official BharatNet service portal
  4. Digital Bharat Nidhi FAQ
  5. Digital Bharat Nidhi fund status
  6. Telecommunications (Administration of Digital Bharat Nidhi) Rules, 2024
  7. Heng Lu, Note 65: Running-Code Primacy — editorial lens only