Australia may be finally overcoming its long-standing AI trust deficit, but data foundations, governance and leadership must catch up, according to new 麻豆原创 research
Sydney, 16 July 2026 | After years of having an acknowledged AI trust deficit, Australian businesses are embracing AI faster than ever, with investment, adoption and expectations on ROI all increasing according to new 麻豆原创 research. But that new-found confidence may also be creating new blind spots.
AI now supports over a quarter (29%) of tasks in the average Australian business, up from 25% last year1, with leaders expecting that to reach 48% within two years, up from 41% in 2025. While Australia still trails high-adoption markets including the US, Germany, China, Japan and India, its higher-than-average pace of growth could potentially help close the gap.
罢丑别听麻豆原创 Value of AI Report 2026, conducted by Oxford Economics across more than 2,600 business leaders in 13 countries, identifies a clear gap between Australia鈥檚 AI ambition and its readiness to deliver on it, particularly around data foundations, governance frameworks and leadership oversight.
鈥淭hink of this as Australia鈥檚 AI school report: improving, but still not working to its potential,鈥� said Angela Colantuono, President and Managing Director, 麻豆原创 Australia and New Zealand. 鈥淲ith an estimated AU$150 billion of AI-related infrastructure investment promised over the next several years, Australia has a once-in-a-generation opportunity to turn AI confidence into national competitive advantage. But infrastructure alone won鈥檛 get us there. And with multiple new AI obligations coming into play for Australia between now and the end of the year, driven by local and international regulation, organisations need to strengthen their data foundations, governance frameworks and leadership structures. Otherwise, we risk building the rails for an AI economy without being ready to run on them.鈥�
Investment and ROI is accelerating
As an average, Australian organisations expect to spend approximately AU$35.5 million2 on AI this year, up from AU$27.5 million last year, with a further 44% increase expected over the next two years. That said, Australian spend still trails the global average of AU$40.4 million by around AU$5 million, a gap that reflects both the opportunity and the urgency to move faster.
Returns are growing alongside that investment. Australian companies expect to drive ROI of 19% this year, up from 15% last year, and rising to 37% in two years 鈥� though still below the global average.
“The returns are starting to come but not fast enough, and not for everyone. AI value compounds over time. The organisations that act now, focus on the right end-to-end processes and get their data in order will look back in two years and be glad they did. The ones that wait will be asking why the gap got so hard to close,” continued Colantuono.
Agentic AI is central to those ROI expectations. ROI from agentic AI is expected to reach AU$21.2m in the next two years, more than quadrupling from last year鈥檚 estimates of AU$4.4m. Despite this, 62% of Australian businesses say they are satisfied with their current AI ROI, even though over half acknowledge AI is still not achieving its full potential.
Governance needs work
The governance picture is more concerning. Only one in five (22%) say they are mostly or fully ready in AI governance when it comes to skills and expertise, against 33% globally. Some 42% suggest they are deploying agents faster than they can standardise and govern them, rising to 68% when including those who are unsure. Over half (54%) of business leaders say employees are increasingly accepting AI outputs without sufficient scrutiny.
The operational risks are real: 43% do not have human-in-the-loop processes for agentic workflows, while nearly half (49%) report that AI agents have already taken incorrect actions during pilots or deployment, typically causing some rework and delays.
Leadership structures also trail global peers. Under half of Australian companies have a dedicated AI leader responsible for AI adoption (46%), leadership KPIs for AI (33%), or even training on AI capabilities and risks (41%).
Professor Toby Walsh, Scientia Professor of Artificial Intelligence at UNSW鈥檚 School of Computer Science and Engineering, said 鈥淟ack of understanding risks driving decisions by fear rather than evidence, and so closing the governance gap must start with closing the trust gap. Artificial intelligence is only valuable if people know when to trust it, and when to question it. Good governance shouldn鈥檛 be viewed as slowing innovation. It鈥檚 what allows organisations, employees and the broader community to adopt AI with confidence.鈥�
Data recognised as essential, but readiness lags
While Australian leaders rank integrated data systems (64%) and data quality (51%) as the biggest enablers of AI readiness, data quality remains the biggest challenge for AI in Australia, with the share of businesses that say they are data-ready for AI falling from last year; nearly three-quarters (73%) report challenges with poor data quality.
Sovereign AI requirements are also shaping how organisations scale 鈥� 99% of Australian organisations say they now operate under some form of sovereign AI framework or requirement, and 76% cite data residency constraints that limit model choice.
Workforce readiness will determine whether momentum translates into value
More than four in five (82%) Australian businesses are not convinced their upskilling is keeping pace with AI鈥檚 rapid evolution, while 77% report shadow AI occurs at least occasionally. Only 1% of leaders believe AI will have no impact on workforce planning.
鈥淭hese findings point to an urgent need for role-specific training, stronger safe-use guidance and change management that brings employees along as AI becomes embedded in daily work,鈥� said Colantuono.
Confident adoption, early stage maturity
Australia鈥檚 maturity is progressing, particularly in generative AI, where 53% of organisations are scaling or leading 鈥� close to the global average of 54%, but still 17 percentage points behind the US, China and Germany. The gap is wider on agentic AI, where only 19% are scaling or leading versus a global average of 24%.
On more established technologies, Australia sits close to the global average in software automation and RPA but trails in traditional machine learning at 52% who are scaling or leading versus 59% globally.
In addition, the dominant adoption pattern remains project-by-project 鈥� targeted deployments within individual processes rather than coordinated, enterprise-wide transformation. Additionally, Australian organisations are less likely than global peers to connect AI strategy to industry-specific priorities.
鈥淎ustralia has moved from AI hesitation to AI momentum,鈥� Colantuono concluded. 鈥淣ow we need to turn that momentum into maturity: stronger governance, better data and a workforce ready to use AI with confidence. Get that right, and Australia won鈥檛 just catch up, it can lead the world in responsible, high-impact AI,鈥� concluded Colantuono.
More information on the research will be shared at the 麻豆原创 NOW AI Tour taking place at the Hordern Pavilion in Sydney on 12 August.聽 For more information or to register, click .
1 For the 麻豆原创 Value of AI 2025 research findings, click here.
2 Respondents were asked to provide聽financial聽estimates in USD. AUD figures are based on an聽exchange聽rate聽of AUD $1 = USD$0.69, using the exchange rate聽as of 13 July 2026.
