Australian businesses are the most responsible AI adopters in the world — and among the least productive. That's the core tension in KPMG's inaugural Global AI Pulse survey, released this week, which surveyed 2,110 C-suite leaders across 20 countries and found Australia leading on AI governance while ranking last among covered markets on productivity gains from automation.
The numbers frame a genuine strategic dilemma for Australian business leaders. Being cautious about AI is defensible. Being cautious while your economy's productivity is going backwards is a different proposition entirely.
What the numbers actually say
The KPMG Global AI Pulse, conducted between 17 February and 17 March 2026, surveyed leaders at organisations with annual revenues of at least US$100 million — three-quarters above US$1 billion. This is enterprise-grade data, not startup sentiment.
On governance, Australia is unambiguously ahead. 31.6% of Australian firms are actively implementing AI governance policies, compared to 26.3% globally. On cybersecurity and data protection, it's even starker: 37.8% of Australian businesses prioritise trust and security, the highest of any market and well above the 25.9% global average.
On productivity, Australia is unambiguously behind. Just 34.7% of firms reported productivity gains from automating workflows — the lowest among all six markets covered — against a global average of 42.3%. Only 38% leverage advanced analytics and real-time insights, below the 41% global mark.
"Australian businesses are much more conscious about trust and the responsible use of AI, compared to businesses around the world, and that's what sets us apart," said KPMG Australia chief technology officer John Munnelly. But he added a blunter assessment: "AI is the biggest lever for productivity change we have. It's up to us to pull it, sensibly, and be trusted in the process."
A pattern, not an outlier
If KPMG were the only survey showing this, you might discount it. It isn't.
Deloitte's 2026 State of AI in the Enterprise report, published in February, surveyed over 3,000 director-to-C-suite leaders and found an even wider gap. Only 65% of Australian organisations plan to increase AI investment in the next financial year, compared to 84% globally. Just 12% of Australian leaders say generative AI is already transforming their business, versus 25% worldwide. And only 28% have moved 40% or more of their AI pilots into production.
"Australian organisations are making strides in AI adoption, but the data shows they are still trailing global peers when it comes to genuine transformation," said Stuart Scotis, Deloitte Australia's national leader for technology, innovation and GenAI. "Leaders need to move beyond incremental improvements and single use cases, and rethink how work is done in a world of abundant digital labour."
Meanwhile, Indeed Hiring Lab's April 2026 report found that two-thirds of all AI-related job postings in Australia come from just 1% of employers — a concentration ratio unchanged in two years. The investment gap isn't just between Australia and the world. It's between a handful of Australian firms moving fast and everyone else standing still.
Why this matters when productivity is already declining
Here's the context that makes this more than a survey story. Australia's multifactor productivity declined 0.5% in 2024-25, according to the Productivity Commission's Annual Bulletin released in February. That's below the 20-year average of 0.4% annual growth and dramatically below the 1.6% average recorded between 1994-95 and 2003-04.
The Productivity Commission has explicitly identified AI as a key lever for reversing this trend, naming "Enable AI's productivity potential" as a priority policy reform area. But when Australian businesses are adopting governance frameworks at world-leading rates while deploying productivity-enhancing automation at world-lagging rates, there's a mismatch between the policy ambition and corporate behaviour.
Globally, the KPMG data shows a clear link between AI maturity and business outcomes. AI leaders — the 11% of organisations scaling agents across functions — report 82% achieving meaningful business value from AI, compared to 62% of their peers. These leaders invest more in talent, deploy agents more deeply across IT, operations, and marketing, and crucially, pair governance with execution.
Australia's approach is the inverse: governance without commensurate execution.
The cultural preference for caution
The KPMG survey includes a revealing data point on how different regions envision human-AI collaboration. East Asian leaders anticipate AI agents leading projects (42%). North American leaders lean toward peer-to-peer collaboration (31%). Australian leaders prefer human-directed AI (34%) — a model where AI assists but humans remain firmly in charge.
That preference aligns with broader cultural patterns. The Australian Responsible AI Index 2025, conducted by Fifth Quadrant and sponsored by the National AI Centre, found a significant confidence-implementation gap: 58% of organisations claim confidence in human oversight of AI, but only 23% have actually implemented it. Similar gaps appear in testing, monitoring, and contestability.
Australia's caution isn't entirely unfounded. The country is building real regulatory infrastructure — new Privacy Act amendments taking effect in December 2026 will require transparency around automated decision-making, with penalties up to AUD $50 million. Anthropic recently signed a Memorandum of Understanding with the Australian government on AI safety, the first deal under the National AI Plan. There's substance behind the governance posture.
But as Deloitte's Dr Elea Wurth, Trustworthy AI Leader for Asia Pacific, put it: "Agentic AI forces Trustworthy AI out of theory and into operation. When systems can decide and act, organisations must embed real risk management — clear accountability, guardrails, and intervention points — into agentic workflows." The argument isn't governance or productivity. It's that you can't have sustainable productivity without governance — and governance without productivity is just expensive compliance.
What to watch
Three things will determine whether Australia's governance lead becomes an advantage or merely a consolation prize.
The December 2026 regulatory deadline. When the Privacy Act's automated decision-making rules take effect, companies with mature governance frameworks will adapt faster than those scrambling from scratch. This is where early investment in responsible AI could pay off — but only for the firms that are actually deploying AI at scale and need those frameworks operationally, not just on paper.
The pilot-to-production conversion rate. Both KPMG and Deloitte flag this as the critical bottleneck. Over half of Australian respondents in the Deloitte survey expect to move 40% or more of AI pilots into production within three to six months. If that materialises, the productivity gap could narrow quickly. If it doesn't, expect the gap to widen.
Whether governance becomes an on-ramp or a barrier. Steve Chase, KPMG's Global Head of AI and Digital Innovation, framed the global challenge precisely: "There is no agentic future without trust and no trust without governance that keeps pace." Australia has the trust. The question is whether it can keep pace.
At Heygentic, we see this tension play out with Australian businesses every week. The organisations getting real value from AI aren't the ones with the most cautious approach or the most aggressive one. They're the ones who built governance foundations and then actually deployed at scale. Australia has the first half right. It's the second half that determines whether responsible AI leadership translates to competitive advantage — or just a slower path to the same destination.
Sources
- Australia leads the world on responsible AI, but lags on productivity gains — KPMG Australia
- Three out of four global leaders will prioritize AI investment despite economic uncertainty — KPMG International
- Australia leads 'responsible' approach to AI adoption — HCA Magazine
- Australian organisations lag global peers in realising AI's transformational potential — Deloitte Australia
- Annual Productivity Bulletin 2026 — Productivity Commission
- Responsible AI Index 2025 — Fifth Quadrant / National AI Centre
