A new report released today from the University of Sydney Business School has created Australia’s first AI Exposure Index for all 150 federal electorates, revealing where AI-driven changes to work could have the greatest political impact.

‘The Wired Belt: Where AI Could Reshape Australian Politics’, produced by Professor Clinton Free and Lachlan Harris, uses occupation-level census data to rank every electorate according to the concentration of occupations potentially exposed to AI-driven automation, augmentation and task redesign.

KEY FINDINGS

• The 30 most AI-exposed electorates are all metropolitan. Not a single rural or provincial seat appears in the top 30 — a reversal of the geography of previous economic shocks.
• They are Australia's youngest, most educated, highest-earning, most culturally diverse and most mortgage-stressed electorates. In the 30 most exposed seats, 44 percent of residents hold a bachelor's degree or above, against 13 percent in the least exposed.
• Labor, the Liberals and the Teals sit on the political front line. Labor holds 78 per cent of the 60 most exposed seats, alongside Teal independents, the Liberals and the Greens — the established urban voting bloc of Australia's major cities.
• AI exposure runs directly opposite to One Nation's support. Cross-referenced against national RedBridge Group polling, the seats most exposed to AI are precisely those least receptive to One Nation today.

Each of these data-backed findings carries significant implications for Australian politics. Taken together, these findings suggest that if AI materially reshapes work, Australian politics could be entering an even more intense period of change, disruption and reinvention.

“For the past two decades, economic disruption in Australia has been largely a regional story: the long decline of farming, manufacturing and carbon-intensive industry,” said Professor Clinton Free from the University of Sydney Business School.

“Any AI disruption to work could reverse that geography. Our index shows its sharpest effects are likely to land on the educated, professional and mortgage-stressed electorates of our major cities — the communities that form Australia's emerging 'wired belt', and much of the established centre of Australian politics.

“The critical point is that this is about job disruption, not just job loss. The political risk doesn't require mass unemployment. It arises if AI quietly erodes the bargain at the heart of these communities: that a good education and a professional career will deliver economic security and socio-economic advancement.”

Lachlan Harris, co-author of the report and former Press Secretary to Prime Minister Kevin Rudd, said: “The data in the report clearly shows that AI disruption is perfectly targeted at the exact seats that are the least likely to swing to One Nation. This means any political moderation provided by urban established party voting is highly exposed to AI job disruption. The political implications of this finding are striking and suggest political change in Australia will only accelerate in coming years.”

The index draws on 2021 Census employment data mapped onto the federal electorate boundaries established by the 2024 redistribution.

The researchers matched AI exposure scores to 474 occupational categories and weighted them according to the actual occupational composition of each electorate.

To reduce reliance on any single methodology, the Index combines three complementary measures:

• Jobs and Skills Australia’s 2025 generative AI automation exposure scores;
• the International Labour Organisation’s global index of occupational exposure to generative AI; and
• the AI Occupational Exposure measure developed by Felten, Raj and Seamans (2021).

The report was prepared with valuable data guidance and analytical assistance from the Australian Parliamentary Library. The full report and detailed rankings for all 150 electorates is available here.

Declaration

The authors acknowledge the generous support of the Parliamentary Library in the preparation of this report.