The Association of Communications and Technology Operators (ACT), representing South Africa's largest telecommunications providers, is seeking a five-year exemption from the Competition Act provisions. The goal of this move is to allow operators to coordinate the implementation of their rural infrastructure.
The proposed idea is to eliminate infrastructure duplication, reduce deployment costs, and cover underserved areas. However, critics consider this idea highly undesirable.
Despite the undeniable need for rural residents to be connected—given the high cost of communication, poor road conditions, unreliable power supply, and difficulties with trunk lines—critics argue that the existence of a problem does not mean the proposed solution is correct, especially when promoted by major players.
Rural residents require guarantees of accessibility, not just the presence of services, and competition is what provides these guarantees. The application, submitted in Government Gazette 55257, goes beyond simple sharing of already built towers or existing resources. ACT proposes that operators provide planning information to an independent third party to help identify gaps and opportunities for coordinated investment.
Even if the application claims it will not affect retail prices or market distribution, it does not alleviate concerns about commercial consequences. In a narrow rural market, the first operator to build a facility may become the only one with a rational reason to be there. If large operators agree on where and when to build infrastructure, who will duplicate those investments for a small group of customers?
Consumer Choice
Consumers do not need to sign exclusivity agreements; economic factors can create exclusivity on their own. This leaves the area resident with only one practical choice. And choice is what sustains network fairness.
Customers switch to other providers when service becomes unreliable, prices are high, or when a competitor offers something better. This threat of attrition forces operators to maintain facilities, improve capacity, and adjust prices. Removing this threat leads to a disappearance of pressure.
What appears as wasteful duplication on an operator's balance sheet might be the only leverage a customer has. Critics point to the opposite experience observed in the fixed broadband segment: as more fiber optic network operators and internet providers entered the market, customers gained more options, and prices dropped sharply due to competition.
Furthermore, the proposal has another flaw: it talks extensively about coverage but does not address the issue of affordability. In many underserved communities, mobile coverage already exists, albeit unevenly. However, a more serious obstacle is often people's inability to afford to use this connection properly.
According to its own study, the Competition Commission, mobile data pricing is structurally 'anti-poor,' as low-volume users pay more per megabyte and are forced to rely on less favorable short-term packages. One recent comparative analysis showed that the cheapest available monthly prepaid package for 5 GB cost 89 rand, while Telkom's package for 49 rand was limited to its own network coverage area.
Smaller providers have demonstrated that another model can work: daily access for approximately 5 rand allows active internet use without a contract or purchasing a large package in advance. This is why video salons still exist in communities with mobile signal: a 3-gigabyte movie is useless if the data cost to download it far exceeds the price of the film itself.
Coverage that people cannot afford is not inclusion; it is a signal on the phone that people are afraid to use.
The proposal also requires excessive trust in promises of future fulfillment. An analysis by Icasa in the report 'ICT Sector Status in 2026' showed that as of October 2025, only 4,377 out of 21,878 state facilities subject to spectrum licensing obligations had been connected—about 20%. These facilities include schools, medical institutions, libraries, and traditional government offices. If existing public obligations remain 80% incomplete, why should the commission believe that another broad authorization granted now and controlled later will yield a different result?
ACT argues that this agreement will promote the participation of small businesses and historically disadvantaged firms. Nevertheless, small wireless internet providers, internet service providers, and independent infrastructure suppliers are not included in the proposed planning scope. Many of them have succeeded for years in marginal markets thanks to lean network approaches and local knowledge.
Stop This
Thus, a process dominated by six large operators could push out precisely the businesses that this exemption is claimed to help. This is not an objection to infrastructure sharing. Co-locating equipment, leasing facilities, roaming, and open access infrastructure all have their place. Sharing what is already built can reduce losses while maintaining independence in decisions about where to compete.
Coordinating future plans between competitors is an entirely different matter, especially if the likely outcome is locking one provider into each marginal market. The Competition Commission must reject the exemption in its current form.
If the commission intends to consider any form of approval, the conditions must be established first, not as a secondary thought: this must include open and non-discriminatory wholesale access, measurable rural buildout targets, affordable retail options, controllable service levels, transparent reporting, and real penalties for non-compliance. Small operators must have a fair path to infrastructure and to the market.
The Competition Act exists because cooperation between competitors can be convenient for participating companies but expensive for everyone else. Rural residents of South Africa should not be forced to give up competition in exchange for an unquantified promise of coverage. They deserve the same as any other customer: affordable services, a working network, and alternatives if the network fails.