On 22 April 2026, the Independent Communications Authority of South Africa (ICASA) confirmed in its 2026/27 Annual Performance Plan that it will open a formal market inquiry into over-the-top (OTT) services, naming Netflix and WhatsApp, to assess their impact on regulated telecoms and broadcasting. ICASA expects to publish a findings document by the end of the 2026/27 financial year and a discussion paper in 2027/28, setting out options that range from keeping today's light-touch approach to imposing new obligations, including financial contributions tied to network usage. That last option revives a familiar and contentious idea: the so-called "fair share" levy.
The strongest case for a contribution
It would be a mistake to wave this away as pure rent-seeking by incumbents. The numbers operators point to are real. ICASA's own State of the ICT Sector report (data to 30 September 2025) shows mobile services revenue falling 7.9% even as national mobile traffic surged 21.5% — a divergence driven largely by OTT substitution, as messaging and streaming displace the SMS, voice and pay-TV products that once funded network build-out. Pay-TV subscribers dropped 9.6% to roughly 6.7 million, and broadcasting revenue fell 4.6%. The Association of Comms and Technology, which speaks for MTN, Vodacom, Telkom, Rain, Cell C and Liquid, argues that a handful of platforms generate most of the traffic, force expensive capacity upgrades, and yet carry none of the universal-service, licensing or local-content obligations that licensed operators bear under the Electronic Communications Act of 2005. Framed that way, asking the largest traffic generators to contribute toward the networks they monetise has an intuitive fairness.
Europe already ran this experiment
The problem is that this is not a new question, and the most rigorous regulator to examine it reached the opposite conclusion. In its preliminary assessment BoR (22) 137, published in October 2022, the Body of European Regulators for Electronic Communications (BEREC) found no evidence that large content and application providers free-ride on telecoms networks. On the contrary, those providers invest heavily in subsea cables, content delivery networks and caching servers placed inside or adjacent to operator networks — infrastructure that actually reduces the cost of carrying their traffic. ISP-commissioned studies put traffic-sensitive network costs at €36–40 billion a year across Europe, but BEREC pointedly noted that such costs never surfaced in telcos' financial statements or loss warnings. The European Commission subsequently declined to mandate network usage fees. BEREC also warned that a "sending-party-network-pays" regime would harm competition, hit mobile virtual network operators (MVNOs), and ultimately raise costs for consumers and businesses.
South Africa would be importing, in 2026, a remedy that Europe's own evidence base could not justify.
Why a network fee misfires
Several features of the South African market make the fee even harder to defend than in Europe.
- Consumers already pay twice over. South Africans buy data; the fastest-growing revenue lines in ICASA's report are fixed data and fibre, with fibre subscriptions up 19.3%. OTT services are precisely what make that data worth buying. A levy that platforms recoup through higher subscription prices, or that operators bank while data revenue keeps climbing, is a transfer dressed up as infrastructure funding.
- Net-neutrality risk. A payment obligation hands operators leverage to throttle, gate or deprioritise traffic from platforms that don't pay — corroding the open-internet principle that has underpinned the sector's growth.
- Asymmetric bargaining power. A bilateral payment regime favours the largest operators and squeezes MVNOs and smaller ISPs, concentrating rather than broadening the market.
Separate the defensible lever from the network tax
ICASA's inquiry should not conflate two distinct questions. There is a legitimate cultural-policy debate — captured in the Draft White Paper on Audio and Audiovisual Media Services and Online Safety (Government Gazette 52972, July 2025) — about whether global streamers should face local-content and investment obligations comparable to those on domestic broadcasters. Reasonable people can disagree on the design, but levelling audiovisual obligations is a coherent policy goal with its own statutory home. A network-infrastructure tax is a different animal: it targets the plumbing, not the content, and rests on a free-riding premise the evidence does not support. Bundling the two would let a weak argument ride on a stronger one.
A proportionate path forward
The encouraging part is ICASA's sequencing. An inquiry and discussion paper before any rule-making is exactly the evidence-first posture a proportionate regulator should adopt — and a marked improvement on jurisdictions that legislated first and gathered evidence later. To make that posture meaningful, ICASA should do what BEREC did: interrogate operators' actual financial statements, identify the real driver of data demand, and ask whether South Africa's connectivity challenge is a free-riding problem or a spectrum-and-affordability problem. The latter is far more likely. With 4G/LTE already reaching 99.5% of the population, the binding constraints on the open internet here are the price of data and the pace of spectrum and infrastructure-sharing reform — not Netflix.
The fair-share idea has surface appeal and a genuine grievance behind it. But the cure on offer would raise consumer prices, threaten net neutrality and tax the very services that make connectivity valuable, all to solve a problem the most careful regulators have been unable to find in the data. ICASA has set itself a sober, evidence-led process. It should let that evidence, rather than the incumbents' framing, decide the outcome.