On 23 March 2026 the Namibian government published a decision in its Gazette refusing a telecommunications and spectrum licence to a satellite internet provider. The Communications Regulatory Authority of Namibia assessed the application against six statutory criteria, including competition, technical and financial capacity, spectrum availability, ownership and control, national security, and compliance history. It failed on ownership, because Section 46 of the Communications Act requires at least 51 percent Namibian ownership absent an exemption and the applicant was wholly foreign-owned. In June the authority dismissed every appeal. Namibia had used the only regulatory instrument that reliably works against a constellation, which is refusal at the door.
Compare South Africa. Its regulator, ICASA, has been investigating allegations that the same service is being provided and used within the country without authorisation, sold through resellers to customers the regulator can only find by looking. It has deployed inspection teams to conduct physical verification alongside other organs of state, and has indicated that if a breach is established, its available responses include lodging a formal complaint with the International Telecommunication Union. Statutory penalties run to R5 million or 10 percent of annual turnover per day. The instrument in Pretoria’s hand is a team of people walking around looking for dishes on roofs.
The two cases describe the same problem from opposite ends. A state can decide whether to admit a satellite operator, and after that decision it possesses almost no independent means of establishing what the operator is doing inside its territory. Every fact that would prove a breach, which account was activated where, when a terminal crossed a border, whether a geofence was applied, what traffic was carried, sits in the systems of the party under investigation. This is not a gap in the rules. It is a gap in the evidence, and no amount of redrafting the licence conditions closes it.
The Evidence Sits With the Party Being Investigated
Terrestrial telecommunications regulation was built on physical facts a state could verify for itself. A tower stands somewhere. A fibre route runs through a right of way. A gateway occupies a building. An interconnection agreement produces call records that a licensed local operator is obliged to retain and produce. Regulatory power rested on the regulator’s ability to observe the network independently of the operator’s cooperation.
A low-earth-orbit constellation removes almost every one of those observable points. What remains inside the jurisdiction is a dish and a subscriber, and the dish reveals nothing about the terms under which it was activated. The control plane that decides whether a given terminal works in a given place is operated abroad, updated remotely, and generates logs held under a foreign legal system. A regulator can seize the hardware. It cannot audit the decision that made the hardware function.
That asymmetry explains why enforcement responses across the continent have been so uneven. It is not that some regulators are stricter and others more permissive. It is that most of them, having no means of proving a violation, are choosing between admitting a provider on terms they cannot verify and refusing one whose service their citizens want.
The state can confiscate a terminal. It cannot inspect, subpoena, or command the network that switched the terminal on.
Ownership Rules Are Doing Work They Were Not Designed For
Both Namibian and South African law impose domestic participation requirements, 51 percent local ownership in the first case and a 30 percent equity share for historically disadvantaged groups in the second. These rules were written as instruments of economic transformation, to move the ownership of national infrastructure toward citizens who had been excluded from it.
They are now functioning as something else: the last available point of jurisdictional attachment. A locally incorporated entity with local shareholders is a body that can be summoned, audited, fined, and held to disclosure obligations. Without it, a government’s counterparty is a foreign corporation whose only presence in the country is a customer list. Equity requirements have quietly become the mechanism through which a state retains anyone to hold responsible, which is a heavy load for a rule that was designed to redistribute ownership rather than to establish jurisdiction.
The Question the Congolese Case Actually Raised
The Democratic Republic of Congo banned the service in March 2024, citing national security and the risk of use by armed groups including M23, then reversed itself and granted a licence in May 2025. The reversal is usually read as connectivity winning over securocrats. The more instructive reading is that the ban was unenforceable in exactly the way that mattered, because terminals move, activation is remote, and a government fighting an insurgency in its east had no independent means of determining which terminals were operating in territory it did not control.
This is the sharpest form of the problem and the one African governments will meet repeatedly. In a conflict, the questions are whether service can be suspended in a defined area, who decides, on what evidence, with what appeal, and whether that suspension can be verified afterwards by anyone other than the provider. Those are wartime powers being exercised through a commercial terms-of-service document. No African state has published a framework for them, and the continent contains several active conflicts in which the answer will be needed before a framework exists.
Who Absorbs the Cost of Enforcing the Rule
Enforcement also carries a domestic political price that falls entirely on the regulator. After Namibia’s refusal, hundreds of citizens formally challenged the decision. The authority that applied the statute became the party accused of denying people connectivity, while the applicant that declined to meet the ownership requirement was cast as the one trying to deliver it.
That dynamic is worth stating plainly because it will decide how many regulators hold the line. An agency enforcing a law against a popular service absorbs the entire cost of doing so, and receives no support from neighbouring states facing the identical question. Fragmented national enforcement against a single global network means each regulator negotiates alone against a counterparty that negotiates continentally, and the weakest terms accepted anywhere become the benchmark everywhere.
The instrument that answers this is not a continental licence, which no African institution has the authority to issue or the capacity to administer. It is a model regional licensing protocol adopted through compatible national rules: terminal registration, auditable geofencing with logs disclosable to the licensing state, continuity obligations for emergency services, a published and reviewable suspension procedure, mandatory reporting of security-related requests, and agreed treatment of cross-border roaming. Every one of those provisions is a demand for evidence rather than a demand for ownership, which is what makes them negotiable.
The verdict is that orbit has not abolished the border. It has changed what a border is made of. For a century a state’s control over communications rested on its ability to see the infrastructure inside its own territory, and that ability has now been removed without being replaced. What is left is a contest over disclosure, conducted one regulator at a time, against a counterparty that holds all the records. Africa is where that contest is being run first and most openly, and the terms settled in Windhoek and Pretoria will be read closely in every other region that connects from above rather than from the ground.
