On 21 April 2026, the Australian Competition and Consumer Commission gave NBN Co written notice that it would take an extra three months — until 2 September 2026 — to finalise the replacement module determination for the regulatory cycle beginning 1 July 2026. The original statutory deadline had been 30 June 2026. It is the kind of procedural footnote that rarely makes headlines, yet it sits at the centre of how Australia actually delivers net neutrality.
Australia regulates the wire, not the conduct
Most net-neutrality debates — the US FCC's on-again-off-again Title II reclassification, the EU's Open Internet Regulation — are arguments about conduct rules: prohibitions on throttling, blocking, and paid prioritisation imposed on retail ISPs. Australia took a different road. By building a single wholesale-only network and forcing every retailer onto the same non-discriminatory access terms, it made the open-internet outcome structural rather than behavioural.
The instrument is NBN Co's Special Access Undertaking (SAU), the binding framework NBN Co gave the ACCC governing wholesale pricing and terms for its open-access Layer 2 services across fibre, HFC, fixed wireless and satellite. The SAU's current regulatory period runs from 1 July 2023 to 30 June 2032, and inside it the network's terms are reset every three-to-five years through a replacement module the ACCC must approve. When every retail service provider buys the same wholesale inputs at the same published prices, no retailer can be structurally favoured — the discrimination that conduct rules police is engineered out upstream.
That is the strongest case for the Australian model, and it is a serious one. A wholesale-only, equivalence-of-access regime delivers a more durable open internet than conduct rules that flip with each change of government. It is worth stating plainly before any criticism: the structural approach has given Australia a decade of pricing certainty that the United States, on its fourth net-neutrality reversal, has never enjoyed.
What the September determination actually decides
The replacement module is where that structural promise meets hard numbers. The ACCC's draft determination, published 31 March 2026, sets the entry-level product, benchmark service standards, allowed expenditure, and the cost of capital for FY2027–29. The headline figures show a regulator pushing hard on NBN Co's spending:
- Capital expenditure of $6.97 billion for 2027–29 — roughly 17.5% below what NBN Co proposed.
- Operating expenditure of $7.89 billion over the same window.
- A weighted average cost of capital of 6.49%–6.56% across the cycle.
- A new entry-level product of 25/10 Mbps, lifting the baseline upload speed from the prior 25/5 tier.
ACCC Chair Gina Cass-Gottlieb framed the draft as protecting consumers "while still allowing the company to invest and operate sustainably," and argued that "higher upload capacity should not be a premium feature, but a baseline expectation." On the entry-level mandate, she is right, and it is proportionate: requiring a modestly faster floor product is a light-touch consumer-protection measure that does not micromanage NBN Co's wider price structure.
Where proportionality gets harder
The expenditure cut is where a pro-investment publication should pause. NBN Co's application sought funding partly to extend its fibre footprint and introduce Low Earth Orbit satellite services — exactly the kind of network upgrade that expands access and future-proofs the asset. Trimming proposed capex by 17.5% and setting a WACC near 6.5% is defensible only if the ACCC's prudency concerns are genuinely about waste rather than a reflexive push to lower headline wholesale costs. A regulator that consistently second-guesses a capital-intensive monopoly's investment plans risks under-building the very network whose openness it is protecting. Open access to a stagnating network is a hollow guarantee.
This is the real tension in structural net neutrality: because there is no competitive pressure to discipline investment, the regulator becomes the de facto investment committee. Get the cost of capital slightly too low, and fibre and LEO rollouts that would have happened simply don't — with no market signal to reveal the loss. That asymmetry argues for the ACCC erring toward NBN Co's investment case where the efficiency evidence is genuinely contested, not against it.
Seen in that light, the three-month extension is welcome, not worrying. The ACCC extended submissions from 28 April to 12 May 2026 and pushed the final determination to 2 September, citing the need to weigh stakeholder responses on expenditure and the cost of capital. A regulator setting the terms of a network until the next cycle should take the time to get the WACC and capex right rather than rush a number it cannot revisit for years. Notably, the SAU locks NBN Co's maximum wholesale price trajectory until 30 June 2032 — so what the replacement module gets wrong on investment and service standards is what it gets wrong, with limited room to correct mid-stream.
The model is sound; the calibration is everything
Australia's structural approach to net neutrality remains the more sensible design — stable, non-discriminatory by construction, and insulated from the partisan whiplash that defines the US fight. But its virtue is also its risk: with competition stripped out, every open-internet guarantee rests on the regulator calibrating a single file correctly. The ACCC has bought itself three more months to do exactly that. The right use of that time is to confirm its capex cuts reflect real inefficiency — and, where the evidence is thin, to let NBN Co invest. An open network that nobody is allowed to under-build is the version of net neutrality worth keeping.