Microsoft survey warns Australian firms lag on AI governance

Microsoft has published the Australian findings from its 2026 Work Trend Index on the use of artificial intelligence at work. The research points to a gap between worker adoption of AI and organisational readiness.

Among Australian AI users, 63% said they were producing work they could not have produced a year earlier, while 68% feared falling behind if they did not adapt quickly. At the same time, only 28% said their organisation's leadership was clearly and consistently aligned on AI strategy and policies.

The figures suggest workplace adoption is moving faster than internal governance, management practices and incentives. Microsoft's survey also found that 51% of Australian AI users felt it was safer to focus on current goals than redesign work with AI, while only 13% said reinvention was rewarded when results were not immediate.

This tension is central to Microsoft's argument that the barriers to wider AI use have shifted. Rather than access to tools, the issue now appears to be whether employers can create the conditions for staff to test, refine and formalise new ways of working.

"This research shows the barrier is no longer the technology - it's whether leaders provide the clarity, culture, and confidence for people to use AI in new ways," said Jane Livesey, President, Microsoft Australia and New Zealand.

"Australians are already racing ahead with AI; the organisations that truly lead will bring their people along, redesign work with purpose, and turn today's momentum into real, lasting outcomes," Livesey said.

Human oversight

The data also suggests workers are not simply handing decisions to software. In the Australian findings, 86% of AI users said they treated AI output as a starting point rather than a final answer and "stay responsible for the thinking."

When asked which human skills become more important as AI and software agents take on more tasks, respondents ranked critical thinking and quality control of AI output equal highest, at 50% each. The results suggest employers may need to place greater weight on review, judgement and accountability as AI use becomes more common in day-to-day work.

Microsoft's analysis of more than 100,000 chats in Microsoft 365 Copilot found that 49% of conversations supported cognitive work such as analysing information, solving problems, evaluating material and thinking creatively. The company argued that AI's role is shifting from routine assistance to support for more complex work.

The Australian data also identified a subgroup Microsoft calls Frontier Professionals, described as the most advanced AI users in its research. Among that cohort, 84% said they were producing work they could not have produced a year ago.

These advanced users were also more likely to report deliberate limits on AI use. Nearly half, 48%, said they intentionally did some work without AI to keep their skills sharp, compared with 34% of other respondents. A larger share also said they paused before starting work to decide what should be done by AI and what should remain with a human, 53% versus 34%.

Manager effect

The research suggests local management behaviour plays a large part in whether AI use matures beyond experimentation. Compared with other respondents, Frontier Professionals were much more likely to say their manager openly used AI, set quality standards for AI-assisted work, created space for experimentation and encouraged more ambitious work redesign.

Team habits also appeared to differ sharply. Frontier Professionals were more likely to say their teams brainstormed and refined business processes together to identify AI opportunities, shared AI tips and mistakes, and discussed quality standards for AI-assisted work. They were also more likely to report that agent workflows, human handoffs and quality standards were documented and repeatable at team, function and organisation level.

Microsoft said global findings showed that organisational factors, including culture, manager support and talent practices, accounted for twice the AI impact of individual effort alone, 67% versus 32%. That framing places the emphasis on internal systems rather than individual enthusiasm.

In Australia, the challenge now appears to be less about persuading workers to try AI and more about turning ad hoc use into structured operating practice. The results imply that organisations that fail to define standards, incentives and decision rights may struggle to capture gains already visible at employee level.

Grant Thornton Australia was cited as one example of an employer trying to formalise AI adoption through staff training and internal programs.

"What resonates in the research is that lasting AI value comes from treating AI adoption as an education and enablement program first and a technology rollout second," said Ben Swindale, Chief Technology Officer, Grant Thornton Australia.

"At Grant Thornton, the real differentiator is our people-first approach, backed with guardrails and always maintaining human oversight for maximum accountability and transparency. Through hackathons, conferences, internal communities, digital credentials, agent building courses, and team workshops, we're empowering our people to find meaningful ways to work with AI, which has helped build momentum across the organisation," Swindale said.

The survey was conducted among 20,000 full-time employed or self-employed knowledge workers who use AI at work across 10 markets, including 2,000 in Australia.