Permission meets the size of India
Published 11 October 2026
A January 29, 2026, parliamentary reply supplied a clear official account of the service-licensing position. DoT had issued Unified Licences with GMPCS authorization to OneWeb India, Jio Satellite Communications and Starlink Satellite Communications. The reply also listed rejected applications from Nelco and Yotta, a withdrawn application from Connect4sure, and Amazon Kuiper's application still under examination. This was direct government confirmation that Starlink held the relevant service authorization. A claim that Starlink had simply been denied a licence would therefore miss a licence the government explicitly said had been issued. pib.gov.in, Press Release Page, 2026-01-29
The reply described satellite communication as suitable for rural, underserved and unserved areas, including coastal, border and mountainous regions difficult to cover through fibre or microwave. It also said licence conditions required providers to serve applicants without discrimination within the service area for the authorized services. That stated obligation joined the public-access purpose to the permission. An operator was not being licensed only to select the easiest or richest locations according to an entirely unrestricted preference. At the same time, the reply did not announce a retail price or a complete rollout. It described the scope of licensed service and the duty attached to it, leaving the actual infrastructure and commercial readiness to be assessed separately. pib.gov.in, Press Release Page, 2026-01-29
Ookla's February 4 report showed Starlink's prominence among satellite-internet tests worldwide. It accounted for 97.1 percent of global satellite Speedtest samples in the third quarter of 2025, with Viasat and HughesNet far behind in that dataset. The share was a share of samples, not a verified share of all satellite customers or revenue. Ookla made a similar qualification when using sample geography to identify leading markets: the results gave relative scale rather than exact customer totals. The dataset offered strong evidence of Starlink's visibility in consumer satellite use, but the boundary of the measurement mattered. It could not by itself establish a monopoly across every form of satellite communication or every customer category. ookla.com, 2025 Global Satellite Broadband Performance Report, 2026-02-04
Performance varied widely across the countries in Ookla's report. Latvia's median Starlink download speed was 187.30 megabits per second in the third quarter of 2025; South Sudan's was 15.87. Both were observations of operating markets, not forecasts for India. The spread made the term Starlink speed insufficient on its own. A location, period and measure were needed. Hardware placement, network conditions and the facilities connecting the space link to the wider internet could all affect the experience. The network's global brand did not imply one uniform result for every user. Its practical value in a new market would depend on the system built there and the alternatives a customer already had. ookla.com, 2025 Global Satellite Broadband Performance Report, 2026-02-04
Latency also depended on more than the satellite's height. Ookla described the journey through the user terminal, satellite, gateways and internet routes, with additional delay possible at each stage. Its report connected large improvements in East Africa to a new Nairobi point of presence, while acknowledging the inference involved. It also recorded Starlink's July 24 outage and the company's explanation that key internal software services in the core network had failed. These were complementary lessons from the operating record. Additional routes and satellites could improve resilience, while common software could still cause a broad interruption. A network could be much less dependent on one terrestrial cable without becoming independent of every shared point of failure. ookla.com, 2025 Global Satellite Broadband Performance Report, 2026-02-04
Ookla's report referred to Starlink's 2025 progress account, which described the manufacturing behind its launch pace. The company said it had exceeded deliveries of seventy satellites a week and expanded its Washington facilities to more than seven hundred thousand square feet. It also reported US production capacity above one hundred and seventy thousand terminal kits per week. These were operator figures, not independent factory audits. They nevertheless described why the network's scale involved a production system on Earth as much as a fleet in orbit. SpaceX was making the user equipment and spacecraft while supplying launches. A rival could design a constellation with comparable ambitions without yet possessing that sequence of production, deployment and service operations. ookla.com, 2025 Global Satellite Broadband Performance Report, 2026-02-04 starlink.com, PROGRESS 2025
The progress account also described collision avoidance and space-weather work. Starlink said it screened orbital traffic before launch, coordinated with other operators and improved the time before newly deployed satellites could assume responsibility for avoidance manoeuvres. It described automated responses to solar storms, which can change atmospheric density and complicate predictions and station keeping. These were claims about its own safety practices. The report did not eliminate the risks attached to a large fleet; it identified work required to manage them. A constellation's operating burden continued after launch and across the lives of its satellites. Capacity growth had to be accompanied by tracking, coordination and the ability to respond when the orbital environment changed. starlink.com, PROGRESS 2025
For re-entry, Starlink described designing components to break up and targeting spacecraft returns over open ocean away from populated islands and heavily used routes. It said successful targeting depended on maintaining attitude control to very low altitude and using variable drag to remove orbital energy. The company acknowledged continuing work to make the method resilient to control failures and atmospheric variation. This was an operator account of the end of a satellite's life, not a guarantee that every return was risk-free. It adds a material dimension to any future sovereign constellation: ownership carries responsibility for safe operation and disposal as well as the benefit of control. Building a fleet means planning what happens when individual spacecraft leave service. starlink.com, PROGRESS 2025
On March 10, the government described the terrestrial work continuing alongside satellite plans. BharatNet had made about 217,805 gram panchayats service-ready by February, according to the parliamentary reply. Under the 4G saturation and other mobile projects, 24,263 towers had been commissioned in previously uncovered rural and remote areas. Service-ready gram panchayats were infrastructure milestones, not a count of households with a working home subscription. Towers likewise represented commissioned facilities rather than a complete measure of each resident's experience. The figures nevertheless showed the scale of the public programme against which satellite service would enter. India was extending fibre and mobile access while considering how satellite networks could address remaining gaps. pib.gov.in, Press Release Page, 2026-03-10
The reply also identified submarine fibre links to the Andaman and Nicobar Islands and Lakshadweep, together with a two-hundred-and-twenty-five-kilometre fibre network within Lakshadweep. It described these links as helping fixed broadband, mobile and other high-speed services expand. The island case was especially useful because distance across water can make a space link appear like the obvious solution. India was also using cables and local fibre in precisely such difficult geography. The relevant comparison was therefore between available routes and their performance, not a rule that remote places must use satellites. A location could gain from more than one form of infrastructure, with the additional route improving both access and resilience. pib.gov.in, Press Release Page, 2026-03-10
The March parliamentary reply included another response to incomplete connectivity: rural-governance systems had been developed with offline capability, multilingual interfaces and accessibility features. That did not replace the goal of internet access. It recognized that public services still had to function while access remained uneven. The approach is useful beside the satellite promise because it identifies a need that cannot wait for every network project to finish. Connectivity can improve delivery, but the design of the service also matters. An online-only process may exclude people at precisely the places a communications programme is meant to reach. The government's account treated the infrastructure gap and the usability of public systems as connected problems with more than one immediate response. pib.gov.in, Press Release Page, 2026-03-10