Set up a National Optical-Fibre Network up to the village level; and Wi-Fi zones in public areas. Deployment of broadband in every village would be a thrust area.

The 2014 manifesto's commitment here is concrete in a way that most infrastructure promises are not. Under next-generation infrastructure it undertakes to "set up a National Optical-Fibre Network up to the village level; and Wi-Fi zones in public areas", and in its rural section it says "Deployment of broadband in every village would be a thrust area."

The project that carries it — BharatNet — predates the promise: it was conceived in October 2011 and run through a special purpose vehicle, Bharat Broadband Network Limited, incorporated in 2012, funded from the Universal Service Obligation Fund. What the promise committed to was finishing it.

On laying fibre, that has substantially happened. The CAG's performance audit, tabled on 12 August 2026, records ₹39,888 crore spent against an approved ₹42,068 crore, and "service ready status in 2.19 lakh GPs (96.70 per cent) against the revised scope of 2.26 lakh GPs" as of November 2025. By any measure of civil works, that is a network built.

The audit's central finding is that a built network is not a working one. "As of November 2025, only 33.20 per cent of service-ready GPs were operational." Two in three panchayats where the fibre had been declared ready were not carrying service. The reason is mostly physical and mostly unfixed: fibre faults or cuts accounted for "approximately 48 per cent of non-operational cases", and the average time to restore a fault "stood at an average of 17 days".

Where the network did work, almost nobody was using it. "Only 17.91 per cent of the available bandwidth was utilised." Major telecom operators — the intended wholesale customers — "withdrew due to service quality concerns", and dark fibre that others wanted to lease went unleased "due to absence of policy or delayed decisions at State Implementing Agency level". The project was designed to be self-sustaining on that revenue; the audit records high maintenance costs against "paltry revenue".

The Wi-Fi half of the manifesto sentence fared worst. In the satellite component, meant precisely for the remote places fibre cannot reach, "BBNL did not install Wi-Fi Access Points even in service-ready GPs, thereby preventing end-users from accessing the intended broadband services."

Planning explains some of the rest. Phase-II was designed for 20 states and reached 13. Detailed project reports — the document that decides where fibre goes — "was carried out on the basis of desktop surveys although physical surveys were mandated", and in Punjab and Bihar the work went ahead with no project report at all. Only six states had reached 100 per cent service-ready status by November 2025, "with delays ranging from two to six years across all the Implementation models". On money, the audit records that BSNL "diverted funds of ₹2,001.43 crore from the project for other purposes", and notes that this has since been recouped.

The audit's own summary is the fairest verdict available: "while BharatNet has achieved substantial infrastructural success, it has fallen short of its broader objective of bridging the digital divide." It reads the 2023 decision to fold BBNL into BSNL and run the project through a single agency as an acknowledgement of exactly the governance and execution problems it documents — and that redesigned programme now aims at 6.4 lakh villages rather than 2.5 lakh panchayats.

Fibre reached the villages. Working broadband reached a third of them. Our verdict is lacks context.