How Australia Uses Claude

How Australia Uses Claude: Findings from the Anthropic Economic Index \ Anthropic Economic Research How Australia Uses Claude: Findings from the Anthropic Economic Index Mar 31, 2026

Anthropic is expanding to Australia. We’re opening a new office in Sydney in the coming weeks, and we’ve signed a Memorandum of Understanding with the Australian government to cooperate on AI safety research and support the goals of Australia’s National AI Plan. To mark the occasion, we thought we’d look more closely into how Australians are using Claude.

Key Findings Australia is among the leading adopters of Claude, accounting for 1.6% of global Claude.ai traffic. Per capita, Australians’ use of Claude is more than four times higher than expected for the size of its population. Adoption within Australia is concentrated in two states: New South Wales (37% of conversations) and Victoria (31%). Per capita Claude usage is lower in every other state and territory. Australia's mix of use-cases generally tracks with its Anglosphere peers: 46% of Claude conversations are for work, 7% for coursework, 47% for personal use. This is a profile typical of high-income, high-adoption economies. But Australia’s mix of tasks is more diverse. Like other countries, Computer & Mathematical tasks remain the single largest category in our dataset—but this category is about 8 percentage points below the global baseline. It’s offset by higher-than-average volumes of office, sales, management, and personal life tasks. Users in Australia tend to prompt Claude for more complex tasks, as measured by the estimated years of schooling required to understand the prompt. At the same time, we estimate that the length of time these tasks would’ve taken without AI is roughly 20% less than average. In other words, relative to how the rest of the world uses Claude, Australians’ tasks are associated with a higher education level but a shorter duration. Like many other high-adoption economies, Australia registers as having a relatively low “AI autonomy” score (3.38 on a 1–5 scale), which suggests that Australians use Claude in more collaborative, less delegated ways.

High adoption overall, unevenly distributed across states and territories Australia accounts for 1.6% of global Claude.ai traffic, ranking eleventh among all countries in our February 2026 sample (Figure 1). Its Anthropic AI Usage Index (AUI) is 4.1, which means Australians use Claude over four times more than its working-age population would predict. This places Australia among the highest per capita adopters of Claude with ranking seventh behind Singapore, Israel, Luxembourg, Switzerland, the United States, and Canada. Figure 1: Top 20 countries by share of global Claude.ai use Australia accounts for 1.6% of global Claude.ai consumer use at rank 11. Bars show each country’s share of the 1M conversations sampled from Claude.ai in February 2026. Australia highlighted in blue.

Within Australia, usage is concentrated in the most populous states. New South Wales accounts for 37.2% of conversations, followed by Victoria at 30.8%, Queensland at 17.7%, and the remaining states and territories at a combined 14% (Figure 2). States and territories’ working age population size explains much of this ordering. However, adjusting for this, New South Wales has an AUI of 1.20 and Victoria an AUI of 1.19. These are the only two states with higher than expected per capita adoption; every other state or territory has an AUI below 1, with adoption lowest in Western Australia at 0.68, Tasmania at 0.32, and the Northern Territory at 0.12. Figure 2: Share of Australia’s Claude.ai use by state and territory New South Wales accounts for the largest share of Australia's Claude.ai use (37.2%), followed by Victoria (30.8%), Queensland (17.7%), Western Australia (7.6%), South Australia (4.6%), the Australian Capital Territory (1.4%), Tasmania (0.6%), and the Northern Territory (0.1%). Data from February 2026. In contrast to the strong association between income and Claude usage per capita we observe across countries , income does not appear to predict adoption across Australian states and territories (Figure 3). That said, since we can only compare across the eight, the lack of correlation between income and usage is suggestive rather than conclusive. Local workforce composition is a likely factor for why adoption appears decoupled from income across Australian states and territories: mining-heavy Western Australia, the state with the highest GSP per capita, has low usage per capita. The Northern Territory likewise has high GSP per capita and low AUI. This likely reflects these sparsely populated states and territories having a relatively smaller share of workers in occupations where Claude tends to be adopted. The Australian Capital Territory has above average income but lower-than-expected Claude usage per capita, which may reflect barriers to adoption among its large public sector workforce. The two states with the greatest adoption, New South Wales and Victoria, have slightly below average income; high adoption in these states likely reflects the higher share of workers in finance, professional services, and tech sectors, where Claude usage tends to be higher. Figure 3: Anthropic AI Usage Index (AUI) and relative gross state product (GSP) per capita by state and territory in Australia Each state's AUI is shown alongside its gross state product per capita, expressed relative to the Australian average ($101,438); both are indexed so that 1.0 represents parity. States and territories are ordered by AUI. Sources: Anthropic Economic Index, February 2026; ABS Estimated Resident Population, June quarter 2025; ABS Australian National Accounts: State Accounts, 2024–25. Australia's use of Claude resembles its Anglosphere peers Our fourth Economic Index report introduced four economic primitives: the use-case mix, the degree of autonomy afforded to Claude, task success, and task complexity. Figure 4 shows where Australia falls on each of these relative to other Anglosphere countries (in orange), other high-adoption economies (in grey), and all countries in our sample (in purple). In terms of how Claude is used, Australia is similar to its Anglosphere peers. 46% of Australian conversations are classified as work-related, close to the median and in the middle of the Anglosphere range. Coursework accounts for 7% of Australian use, which is toward the bottom of the distribution (and…