How do Australians feel about AI?
Australians’ wariness of artificial intelligence (AI) has become entrenched.
Australians’ wariness of artificial intelligence (AI) has become entrenched.
Subscribe to The Informer for monthly expert analysis, and to Events for advance notice of visiting world leaders and distinguished guests.
You may unsubscribe from Lowy Institute newsletters at any time. For information on our privacy practices and how to unsubscribe, see our Privacy Policy.
How do Australians feel about AI?
A recent mini-poll commissioned by the Lowy Institute finds that Australians’ wariness of artificial intelligence (AI) has become entrenched. Conducted in October, amid the technology’s rapid diffusion and prominence in the national debate, the snapshot finds that 65% of Australians feel the risks of AI outweigh the benefits, while 33% feel the opposite. The result is unchanged from the Institute’s March poll, but represents a 13-point increase from 2024, when both the technology and discussion around AI were less mature (see Figure 1).
Australians stand out internationally in their concerns about the technology itself. In a Pew Research Center survey of 37 countries conducted earlier this year, Australians were among the most likely of any nationality to say the growing use of AI in daily life left them more concerned than excited. Job losses are a particular worry: 76% of Australians said AI would lead to fewer jobs in Australia over the next 20 years — the joint-highest share of any country surveyed, and well above the median of 46%. Just over half (52%) also expected AI to widen the gap between rich and poor, second only to South Korea.
These concerns sit against a gloomy economic backdrop. In the Institute’s March poll, six in ten Australians were pessimistic about the country’s economic performance over the next five years — a record high for the Lowy Institute Poll, and gloomier than during the Global Financial Crisis or the Covid pandemic. Little has happened since to lift the mood: inflation remains above target, and the Reserve Bank has raised interest rates three times since March. The AI boom is itself part of the picture, with the Reserve Bank flagging that AI-related investment could push up input costs and add to capacity pressures.
Australians’ anxiety about AI has a geopolitical element too. Only 23% of Australians say they feel “very comfortable” or “somewhat comfortable” with Australian businesses using AI sourced from China. American AI fares only slightly better. Just 36% say the same about AI sourced from the United States (see Figure 2).
Some of this discomfort reflects how Australians see the two superpowers. In the Institute’s March poll, only 31% of Australians trusted the United States to “act responsibly in the world”, and 28% said the same of China (see Figure 3).
Yet near-parity in trust does not carry over to technology. Comfort with American AI exceeds comfort with Chinese AI by 13 points, more than four times the gap in trust. Australians appear to see something specific in Chinese technology that worries them, and they have reason to. China’s National Intelligence Law requires Chinese companies, including AI developers like DeepSeek and Zhipu, to support and cooperate with state intelligence agencies when asked. The Australian public is alert to the wider risk. In March, 62% described “cyberattacks from other countries” as a critical threat to Australia’s interests, and a similar share expected China to pose a military threat in the next 20 years.
This points to a tension in the Australian government’s agenda. Its National AI Plan, released in December 2025, aims to make Australia a leading destination for data centre investment, and in March the government set expectations for developers on the national interest, energy, water, jobs, and innovation. The ambition rests on Australia’s abundant land and energy, and on the logic of the alliance: computing power hosted in Australia, largely by American firms, keeps the region’s AI future tied to the United States rather than China. That logic is the strategy’s strongest public foundation, with 73% of Australians in the Institute’s March poll saying the alliance was important to their security. The economic case is substantial too. The Productivity Commission has estimated that AI could add more than AU$116 billion to Australian economic activity over the next decade.
An emerging risk is that government policy outpaces public sentiment. The strategy asks Australians to host infrastructure for a technology whose risks most of them believe outweigh its benefits, built largely by American firms at a time when only a third of Australians trust the United States. The gains also remain prospective while the costs are already visible. The Reserve Bank has cut its productivity forecast for 2026 even as data centre construction adds to price pressures. The United States offers a warning. At the November midterm elections, voters in dozens of local ballot measures almost uniformly backed limits on data centres, according to one election-monitoring project. Yet Australia’s build-out is accelerating, with data centre companies having raised at least AU$35 billion so far in 2026, up 46% on 2025, according to Reserve Bank analysis.
Australians’ support for the alliance does not currently extend to American AI. That may change, as public sentiment can swing fast. But as with any major policy agenda, the government may need to reckon with an emerging trade-off between Australian public sentiment and the country’s geostrategic imperatives on AI.
The 2026 Lowy Institute mini-poll on artificial intelligence reports the results of a nationally representative survey of 3,028 Australian residents aged 18 and above, conducted from 6–18 October 2026 by the Social Research Centre (SRC) on behalf of the Lowy Institute. The survey uses the Life in Australia™ panel, which is the only probability-based online panel in Australia.
On a simple random sample of 3,028 responses, the margin of error is approximately 1.2%. Where a complex sample is used, the ‘design effect’ measures the additional variance in comparison with a simple random sample. The design effect for this survey is 1.21. For the 2026 Lowy Institute mini-poll on artificial intelligence survey, a completion rate of 74.8% was achieved. Taking into account the recruitment rate (6.1%) into the panel and attrition from the panel, the cumulative response rate is 2.2%.
SRC uses a mixed-mode approach for the survey, including online surveys (99% of respondents) and computer-assisted telephone interviewing (1% of respondents) to provide coverage of the offline population (households without internet access).
About the author
Charles Lyons-Jones
Charles Lyons-Jones is a Research Fellow in the Lowy Institute’s Foreign Policy and Public Opinion Program. He oversees the annual Lowy Institute Poll and the Global Diplomacy Index.
Subscribe to The Informer for monthly expert analysis, and to Events for advance notice of visiting world leaders and distinguished guests.
You may unsubscribe from Lowy Institute newsletters at any time. For information on our privacy practices and how to unsubscribe, see our Privacy Policy.