Anthropic published new data showing Australia accounts for 1.6% of global Claude.ai traffic, ranking 11th worldwide despite representing just 0.3% of global population.

The AI safety company's Economic Index found Australians use Claude over four times more per capita than expected, placing Australia seventh globally behind Singapore, Israel, Luxembourg, Switzerland, the United States, and Canada.

Usage within Australia concentrates heavily in two states. New South Wales accounts for 37% of conversations and Victoria 31%, with both states showing above-average per capita adoption. Every other state and territory falls below the national average, with Western Australia at 0.68x expected usage and Tasmania at just 0.32x.

Task patterns mirror Anglosphere peers

Australian Claude usage splits 46% work-related, 47% personal, and 7% coursework — a profile typical of high-income economies. This contrasts with lower-adoption countries where coursework represents two to three times the share.

However, Australia shows a more diverse task mix than the global baseline. Computer and mathematical tasks remain the largest category but sit 8 percentage points below the global average, offset by higher volumes of office, sales, management, and personal life tasks.

Australians tend to prompt Claude for more complex tasks requiring higher education levels but shorter completion times — roughly 20% less time than the global average without AI assistance.

The country registers a relatively low "AI autonomy" score of 3.38 on a 1-5 scale, suggesting Australians use Claude collaboratively rather than delegating tasks entirely.

Geographic concentration reflects workforce composition

Income doesn't predict Claude adoption across Australian states, unlike the strong correlation Anthropic observes between countries. Mining-heavy Western Australia has the highest gross state product per capita but low Claude usage, likely reflecting fewer workers in occupations where AI tools are commonly adopted.

The Australian Capital Territory shows below-expected usage despite above-average income, potentially reflecting barriers to adoption among its large public sector workforce.

Anthropic released the findings as it expands into Australia with a new Sydney office and a memorandum of understanding with the Australian government on AI safety research supporting the National AI Plan.