The mechanism nobody is naming
Here is the thing that makes the immigration debate different in 2026 from every version of it before: the shortage that justified the intake is being absorbed by software, not by the people brought in to fill it.
Large-language-model deployment does not eat manufacturing jobs first. It eats graduate-entry and entry-level knowledge work first: document processing, customer triage, compliance checking, data entry, basic coding, research synthesis. These are the roles that a newly arrived student or recent graduate competes for. They are also, by construction, the roles with the shortest remaining runway before a model does them faster and cheaper.
Now look at where the intake is concentrated. Of the 636,450 foreign nationals who arrived in Australia in 2023 under the ABS 12-in-16-month residency measure, 282,570 arrived on study visas. That is 44.4% of the foreign-national intake in a single visa category, concentrated in exactly the occupations that large-language-model deployment absorbs first.[^1]
The composition of net migration tells the same story from a different angle. Students made up 51.7% of net overseas migration in 2023. Skilled migrants, the category the intake is nominally justified by, were 14.1% combined, with permanent Skill alone at 5.8%.[^1] The intake is not primarily a skilled-worker program. It is primarily a student program, and the study pathway leads into the graduate labour market that automation is entering fastest.
This is not a forecast dressed up as a fact. The displacement of entry-level knowledge work by language models is documented in employer statements and filings and is already visible in graduate hiring. The argument is that the intake was sized for a labour shortage that is being resolved by a different mechanism, and that continuing to run the intake at its current composition while that mechanism accelerates is a structural cost, not a solution.
What the numbers actually show
The 2023 figures need to be read carefully. Net overseas migration peaked at 530,700 in 2023, a post-border-reopening catch-up, and has since fallen to 330,500 in 2024 and 300,800 in 2025, a 43% reduction from the peak.[^3] Any argument that the intake is still climbing is factually wrong, and this piece does not make it.
What the composition argument does not need is a rising total. The student-stream dominance held in 2019, held in 2022, and held in 2023. The mechanism is structural, not a product of one exceptional year.
The economic effect of running a high-intake model is visible in the national accounts. GDP rose while GDP per capita fell in 7 of the last 12 quarters, and in 25 of 104 quarters since 2000, through to 2026-Q1.[^2] The ABS national accounts measure this in chain volume terms, seasonally adjusted, quarter on quarter. Output per person went backwards while headline GDP rose. Population growth is one input among several, including hours worked, terms of trade and investment, and this piece does not claim it as the sole cause. What it does claim is that a growth model that adds people faster than it raises output per person is not building wealth for the people already here. It is diluting it.
The living cost indexes make the same point from the household side. Between 2012-Q1 and 2026-Q2, living costs for other government transfer recipient households rose 51.7%, for pensioner and beneficiary households 50.0%, and for age pensioner households 48.5%.[^3a] These are the ABS Selected Living Cost Indexes, which include mortgage interest charges and move differently from the CPI. They are not what any individual household experienced. They are what the index for that household type experienced, and the direction is unambiguous.
The student pathway and the system that runs it
The Commonwealth's own Rapid Review into the Exploitation of Australia's Visa System, the Nixon Review, documented how the student pathway has been used for purposes other than study: non-genuine students, visa hopping onshore, and the providers and agents who facilitate it.[^5] This is a finding about the system and about named providers. It is not a claim about students as a group.
The settings that permit onshore visa switching from a study visa into work are a policy choice, not a natural law. A government that wanted to tighten them could. The Nixon Review gave it the documented basis to do so. The question is whether any government will.
Why One Nation leads rather than follows
The Coalition presided over high intakes during its own time in government. Its current position on reducing migration is a change of mind rather than a demonstrated conviction. That is not an attack. It is a factual description of the record, and it is the precise reason the coalition thesis has One Nation leading rather than following.
The Liberals and Nationals bring the parliamentary numbers and the ministerial experience. What neither has demonstrated in government is the willingness to cut the intake hard and hold it there when the business lobby pushes back. One Nation's proposition is not a trim at the margin. It is taking the intake down hard and continuing to take it down, year on year, paced against measured automation displacement.[^6]
A schedule can be modelled, costed and defended on air. "We would reduce it" cannot. The continuing nature of the reduction is the point. One cut that is then allowed to drift back up is the pattern both major parties have already run. A year-on-year commitment with a stated mechanism for pacing it is a different instrument entirely.
What the reduction buys
The benefit case is concrete and it leads, because a vision converts where a warning only alarms.
Housing completions that outpace household formation rather than chase it. When the intake runs at 530,000 net, completions cannot keep up and the deposit keeps moving. When the intake falls and holds lower, the ratio shifts. This is a mechanism, not a promise.
Infrastructure spend per head rising instead of being diluted. Every dollar of Commonwealth infrastructure investment is divided across a larger population when the intake is high. A lower intake means the same dollar goes further for the people already here.
Wage growth in the occupations most exposed to both automation and labour competition. Entry-level knowledge workers and recent graduates are the cohort that faces both pressures simultaneously. A lower intake in the study stream reduces one of them.
Public services sized for the population that is actually here. The argument about the dependency ratio and ageing is real, and it deserves a direct answer. The Productivity Commission's inquiry into migrant intake found that migration "does not offer a long-term panacea" because "immigrants age too."[^4] The dependency ratio does not stay fixed when you import workers to manage it. It shifts forward. The structural answer is raising productivity of the existing population, not perpetually expanding the denominator.
Australia holds the energy, food and commodity base to be among the richest countries on earth per head. Norway turned oil into roughly US$1.8 trillion in sovereign wealth for 5.5 million people. Australia exports more resource value and has no equivalent fund. That is a promise, not a grievance. Used domestically, cheap energy makes Australian-owned compute viable and generates the revenue to invest in the population already here rather than carrying debt forward onto them. The migration reduction is one part of a larger reorientation toward output per person rather than output by headcount.
The permanent residency pipeline
One further mechanism that the headline number does not capture: a permanent residency grant is not a one-off. It creates a future entitlement to sponsor family, and the migration program runs a separate Family stream with its own planning level.[^7] Student visas carry dependants. The standing population effect of a year's intake exceeds that year's headline number, and a government that plans only to the headline is planning to the wrong figure. This piece does not pretend to know the exact multiple, because an invented multiplier is the easiest thing for an opponent to disprove. The mechanism is real and citable. The point is that the real number is higher than the headline, and any serious reduction policy needs to account for the pipeline, not just the annual flow.
The technology policy gap
One Nation's published policy index covers 29 policies.[^6] None of them address AI, automation, data infrastructure or telecommunications. This is not a gotcha. It is a structural fact about a party that is about to become a governing force as the data centre build-out accelerates across Australia.
The foreign ownership and energy sovereignty positions One Nation already holds are exactly the right foundations for a technology policy. Who owns the compute, who owns the data centres, what the Commonwealth's AI procurement stack costs and where the money goes: these are questions that flow directly from existing One Nation positions on foreign ownership and domestic energy. The section will argue what that policy should look like as the build-out continues, because the positions are there and the policy is not yet written. That is a genuine opportunity, not a weakness, if the party moves to fill it.