On 12 May 2026, Bangladesh's Posts and Telecommunications Division — acting on a proposal from the Bangladesh Telecommunication Regulatory Commission (BTRC) — granted Banglalink, the local arm of VEON, a two-month Proof-of-Concept authorization to trial Starlink's satellite Direct-to-Cell (D2C) service. The trial targets exactly the places terrestrial towers never reach: the hill tracts, the chars, coastal islands, and up to 22 nautical miles into the Bay of Bengal, where deep-sea fishermen have historically gone dark. Banglalink will run the service over its own 1920–1925 / 2110–2115 MHz spectrum, initially limited to OTT-based messaging, with the approval explicitly described as non-precedential for commercial operations.
The connectivity case is strong and worth stating plainly. Bangladesh has 100% nominal 3G/4G footprint and roughly 185 million mobile connections, yet close to half the population — on the order of 87 million people — remains effectively offline, and rural users are markedly less likely to get online than urban ones. Satellite D2C does not fix affordability or literacy, but it closes the one gap terrestrial economics never will: the last islands, sandbars, and open water where no tower will ever pay for itself. VEON's framework agreement with Starlink Mobile, announced 22 April 2026, leans on real precedent — Kyivstar in Ukraine reported more than five million unique customers connecting via Starlink satellites since its November 2025 launch, and Beeline Kazakhstan completed Central Asia's first satellite-powered WhatsApp call in December 2025. A 'cell tower in space' reaching a standard 4G handset is no longer a demo; it is a roaming partner.
The condition that should give readers pause
The trial's terms require full compliance with Bangladesh's lawful-interception framework, technical capability to hand law-enforcement the data it requests, and BTRC online and offline monitoring of traffic, signalling, authentication, and spectrum use. That mirrors the NGSO satellite guidelines BTRC issued on 26 March 2025, whose Article 26(4) obliges operators to route activity through a local gateway and grant a 'designated national agency' access for interception.
Steelman the rule first, because it is not frivolous. A satellite link that bypasses domestic infrastructure is a genuine national-security and law-enforcement concern: it can carry trafficking, militant coordination, or sanctions-busting traffic that no local provider can see. Every serious jurisdiction — including the EU and the US — maintains lawful-access regimes for telecoms. Requiring a gateway and a documented interception capability before live satellite traffic crosses Bangladeshi territory is, in the abstract, a legitimate exercise of regulatory authority over spectrum the state controls.
The problem is not interception in principle. It is the legal architecture this capability plugs into.
A surveillance regime that has not earned the trust
Bangladesh's lawful-access powers sit atop a statutory lineage with a documented record of abuse. The Digital Security Act 2018 was used, by ARTICLE 19 and CGS tallies, in over a thousand cases between 2018 and 2022, hundreds of them against journalists and politicians; Reporters Without Borders called it one of the world's most draconian laws for the press. Its 2023 successor, the Cyber Security Act, was condemned by Amnesty International as a 'replication' of the same draconian core, and was invoked against critics during the 2024 protests. After Sheikh Hasina's fall, the interim government repealed the CSA and gazetted the Cyber Security Ordinance, 2025 (Ordinance No. 25) on 21 May 2025 — dropping nine of the most contentious speech-criminalizing sections, a real improvement. But the interception plumbing was never rebuilt to match. ARTICLE 19 warned that the accompanying Telecommunication (Amendment) Ordinance 2025 preserved broad surveillance powers without independent judicial authorization, and even a Dhaka University analysis cited by Prothom Alo flagged that public distrust of wiretapping 'stems from a lack of transparency.'
This is the proportionality gap. The Starlink trial bolts a powerful new interception surface — direct-to-handset traffic from the country's most remote and least-monitored citizens — onto an oversight framework that still lacks a warrant requirement, published statistics, or an independent reviewer. The technical mandate is precise; the safeguards around it are not.
What proportionate regulation would look like
None of this argues against the trial. It argues for finishing the reform the interim government started. Three fixes would let Bangladesh keep the connectivity win without expanding unaccountable surveillance:
- Judicial authorization on the record. Interception of D2C traffic should require a specific, time-bound judicial order — not a standing administrative demand to a 'designated agency.'
- Transparency reporting. BTRC and the operator should publish aggregate interception-request counts, as mature regimes do, so that monitoring can be audited rather than assumed.
- A genuine sunset and scope-lock. The approval is already non-precedential and time-boxed; the data-handling and interception terms should be too, with traffic data deleted at the PoC's end absent a court order.
Bangladesh is making the right call on access — extending the digital frontier to the people the market left behind is good policy and good for the tech sector. But connectivity granted through a surveillance pipe is connectivity on a leash. The country has spent two years unwinding the worst of its digital-speech laws. It should not let a genuinely promising satellite trial quietly re-entrench the interception architecture those same reforms were meant to fix.