Australia’s AI ambitions expose the folly of short-term broadband policy

Government should judge the NBN by its contribution to productivity, education, healthcare and participation, alongside its financial performance, writes Paul Budde.

FOR YEARS, I have argued that broadband should be planned as national infrastructure, with a horizon measured in decades. Australia’s costly detour through copper-based NBN connections showed what happens when that horizon is shortened to fit political promises.

The OECD’s latest broadband statistics, released on 13 October, strengthen that argument.

Across member countries, gigabit subscriptions increased from 4% of fixed broadband subscriptions in 2020 to 24% in 2025. Fibre’s share rose from 31% to 51%.

The OECD identifies AI, cloud computing and connected devices as drivers of growing connectivity requirements.

The lesson for Australia is clear: we should build capacity ahead of demand and make it accessible to the people and businesses expected to use it.

That was the logic behind the NBN's original fibre vision.

The Coalition sold its decision to rely more heavily on existing copper as a faster, cheaper approach. Its 2014 policy announcement set objectives of at least 25 megabits per second for all premises and 50 Mbps for 90% of fixed-line premises.

In my view, this confused an immediate service target with a national infrastructure strategy. As I argued in Fibre after the fact: NBN’s retrofit scramble confirms flawed copper legacy; the subsequent retrofit exposed the consequences of that choice.

NBN’s success vindicates original vision — but the job's not done

NBN Co's FY26 results demonstrate the transition from copper to fibre is producing the benefits many predicted from the beginning.

Australia is now investing billions to replace copper.

The final fibre-to-the-node upgrade program, announced in January 2025, is backed by up to $3 billion in government equity and more than $800 million from NBN Co. It covers approximately 622,000 premises, with more than 95% expected to have a full-fibre upgrade option by 2030.

I welcome this investment. But we should remember why it is necessary when the next argument emerges that Australia can postpone infrastructure because existing services are “good enough”.

There is encouraging progress.

NBN Co’s FY26 results report more than 575,000 copper-to-fibre upgrades during the year, alongside fewer network incidents and lower operating expenditure.

By June 2026, 37% of services were on tiers of 500 Mbps or above, compared with 4% a year earlier.

That shift partly reflects the acceleration of existing speed tiers. It should not be presented as proof that everyone independently decided to buy faster broadband. Nevertheless, the network is becoming more capable.

The next task is to turn that capability into broadly shared benefits.

NBN delivering but bottlenecks at home hamper performance

The NBN is finally delivering — but the bottleneck lies with ageing equipment inside the home.

In OECD slams broadband inequality in Australia, I argued that coverage must be accompanied by affordability and digital skills. The same applies to AI.

A regional business needs a reliable service and the capability to use new tools. A low-income household needs an affordable connection and suitable equipment.

We should also avoid using AI as a sales pitch for unnecessary upgrades. Many everyday AI applications work perfectly well without gigabit broadband.

The investment case rests on the combined requirements of cloud services, video, business applications and growing numbers of investors.

In AI is re-wiring the network: Why Australia must rethink its digital and energy future, I mentioned that Australia needs to integrate its telecommunications, computing and energy strategies. Digital skills and inclusion must be part of that planning.

Government should judge the NBN by its contribution to productivity, education, healthcare and participation, alongside its financial performance.

Affordable services, reliable connections and support for effective use should be explicit policy objectives.

The OECD figures offer another reason to sustain the return to fibre. Having spent years correcting a short-term infrastructure decision, Australia should now plan for the decades ahead — and ensure that the resulting opportunities are shared.

Paul Budde is an IA columnist and managing director of independent telecommunications research and consultancy, Paul Budde Consulting. You can follow Paul on Twitter @PaulBudde.