Australians have embraced artificial intelligence faster than they have learned to trust it. Five charts on a nation of reluctant users.
Australia has, almost without noticing, become a nation of artificial-intelligence users. Every few months or more, we reach for an AI tool. Yet only a third of us trust them — lower than the global average of 46%, and coupled with exceptionally high concern. We have been swept up by the technology before coming to terms with it. These five charts illustrate that troubled relationship: what Australians are doing with AI, how we shape up against the rest of the world, where trust is heading, and what it is quietly costing us along the way.
Data: Gillespie, N., Lockey, S., Ward, T., Macdade, A., & Hassed, G. (2025). Trust, attitudes and use of artificial intelligence: A global study 2025 (The University of Melbourne & KPMG). 48,340 respondents across 47 countries.
Australians use AI, fear it, and want it reined in — yet only 36% trust it
Share of Australian respondents. Hover any point for detail.
Source: Gillespie et al. (2025), Trust, attitudes and use of AI: A global study 2025 — Australia insights.
The conundrum is right here. More than three-quarters of Australians (78%) are concerned about AI, while fewer than a third (36%) trust it — and more than 75% want it regulated. This is not the techno-optimism of the past. This is adoption born of caution.
Trust in AI is a rich-world problem
Each dot is a country; dashed lines mark the global averages. Click a legend entry to filter, or drag to zoom.
Hover a dot for its country, use and trust. Click “Advanced” or “Emerging” in the legend to isolate one group.
Source: Gillespie et al. (2025), Figures 2 and 13.
The countries with the highest trust in AI are emerging economies — India, Nigeria, China, Egypt — grouped in the top-right. Advanced economies generally sit lower. Australia (red) lands in the bottom-left “wary” quadrant: below-average use and below-average trust. The optimism is largely in economies where AI is seen as a ladder up.
As AI use spread, trust fell and worry grew
Across the 17 countries surveyed both before ChatGPT (2022) and in 2024. Hover a year to compare all three measures at once.
Willingness to rely on AI and perceived trustworthiness both fell; worry rose.
Source: Gillespie et al. (2025), Figure 16 (17 countries surveyed in 2022 and 2024).
Familiarity has not bred confidence. Even as the tools became widespread, our willingness to rely on AI dropped from 52% to 43%, while worry about it rose from 49% to 62% between late 2022 (just before ChatGPT) and late 2024. The more accustomed we have become to using AI, the more wary we have become of it.
Australians are more sceptical than the world
Below the global average on trust, use and benefits — but keener on regulation.
Red = Australia, navy = global average. Hover any point for the exact value.
Source: Gillespie et al. (2025), global report and Australia insights.
Australians are less trusting than the rest of the world, use AI less, and expect or feel fewer of its benefits. The one measure where we run ahead is the appetite for regulation (77% vs 70%) — we want guardrails first.
Low trust hasn’t stopped risky use
Share of Australian workers who use AI reporting each behaviour.
Among Australian employees who use AI at work.
Source: Gillespie et al. (2025), Australia insights — workplace indicators.
This is the sting in the paradox. The way Australians actually use AI is far from cautious: most have relied on it without checking the output, three in five have made mistakes because of it, and half have passed its work off as their own. We don’t trust AI — and we lean on it carelessly anyway. The fix is less about better algorithms than about AI literacy: knowing when to rely on the machine, and when not to.
The bottom line: Australia’s challenge isn’t getting people to use AI — half of us already do. It’s closing the gap between how heavily we rely on it and how little we trust it. That is a job for AI literacy, not better algorithms.
Gillespie, N., Lockey, S., Ward, T., Macdade, A., & Hassed, G. (2025). Trust, attitudes and use of artificial intelligence: A global study 2025. The University of Melbourne & KPMG. https://doi.org/10.26188/28822919
KPMG, & The University of Melbourne. (2025). Trust, attitudes and use of artificial intelligence: A global study 2025 — Australia insights. https://kpmg.com/au/en/insights/artificial-intelligence-ai/trust-in-ai-global-insights-2025.html
R Core Team. (2025). R: A language and environment for statistical computing. R Foundation for Statistical Computing. https://www.R-project.org/
Country-level figures (Charts 2–3) were read from the published report’s figures (Figures 2, 13 and 16). Verify a sample against the report before final submission.