70% of Australian organisations admit AI outpaced their data control
Fujitsu research finds 70% of Australian organisations share data faster than they can govern it, with self-rated control performance behind the US, UK and Japan.

Why it matters
- 70% of Australian organisations say AI pushes them to share data beyond their governance capabilities, vs 60% in the US (Fujitsu, 17 Nov 2026).
- Only 44% of Australian organisations have governance for how AI systems learn from and retain data after deployment, vs 54% in the US and 45% in the UK.
- 69% of Australian businesses cite a lack of data and AI governance skills as a major challenge, vs 56% in the US.
- 47% of Australian organisations rate their own data and AI control performance as 'good', the lowest alongside Japan, behind the US (50%) and UK (58%).
- 57% of Australian organisations report data sovereignty has become a more prominent board-level topic.
Seven in 10 Australian organisations admit that AI is pushing them to share data more widely than their governance capabilities can safely support, according to new global research from digital transformation company Fujitsu, published 17 November 2026. The figure is 70% locally, against 60% in the United States. The same technology that promises to transform business is, by Fujitsu's account, simultaneously undermining organisations' ability to control their data.
The numbers matter beyond Australia. They quantify a structural problem of the AI deployment cycle: governance frameworks designed for a pre-AI world were never built for the pace or complexity that machine learning systems now impose on data flows. The result, the research argues, is a widening gap between what organisations share and what they can safely manage — a gap that boards, regulators and customers are beginning to notice.
How far behind is Australia?
The research charts a growing disconnect between how data moves through organisations and how effectively they track and manage it. On nearly every metric measured, Australia trails its AUKUS partners.
- Secure data sharing: 69% of Australian organisations report that enabling secure data sharing across ecosystems is a major challenge, compared with 49% of US businesses and 58% of UK businesses. The consequence, according to the research, is a lack of visibility into where critical data flows, where it persists, and who ultimately has access to it.
- Post-deployment governance: Fewer than half of Australian organisations (44%) have clear governance in place to control how AI systems continue to learn from, reuse or retain data after they go live. The US sits at 54%, the UK at 45%. Once an AI model is deployed, many organisations simply lose sight of what happens next — what the research calls a failure of long-term data custody.
- Skills: 69% of Australian businesses identify a lack of skills and expertise in data and AI governance as a major challenge, level with the UK (69%) and well above the US, where only 56% see skills as a major hurdle.
- Self-assessed performance: Fewer than half (47%) of Australian organisations rate their own performance as "good" at maintaining effective control and governance over data and AI systems. That places Australia at the bottom of the pack alongside Japan, behind both the US (50%) and the UK (58%).
The pattern is consistent. Australian organisations know they have a problem, and by their own assessment they are not solving it.
Why does accountability without visibility matter?
What makes these gaps a boardroom issue rather than an IT issue is attention from the top. Data sovereignty has become a more prominent topic in board-level discussions for the majority (57%) of Australian organisations, the research finds.
But directors are overseeing systems they can no longer fully control. The research describes this as "accountability without visibility" — executives held responsible for data outcomes they cannot adequately monitor or influence. For any board, the analysis argues, this is an untenable position.
The external pressure is intensifying on three fronts identified in the research: regulatory scrutiny is growing, customer expectations around data handling are rising, and reputational consequences for missteps are severe. Yet the operational machinery beneath the boardroom remains opaque. When autonomous agents — AI systems that act on data with limited human intervention — start interacting with customer data without guardrails, the research warns, a board cannot afford to hide behind compliance checklists.
What should organisations do?
Fujitsu's prescription is a shift in mindset from static "data protection" policies to dynamic "data control" capabilities, structured around three moves.
1. Extend governance beyond go-live. The moment an AI model moves into production should not be the end of governance; it should be the beginning of a new phase. Organisations need technology that provides continuous visibility into how AI models use, retain and learn from data throughout their operational life.
2. Invest in people. The skills gap will not close itself. The research calls on Australian organisations to prioritise building internal capability in data and AI governance, and critically, to establish clear accountability that extends beyond the IT department. Data sovereignty is a business-wide responsibility that requires business-wide ownership.
3. Shift from policy to proof. Organisations need dynamic architectures that can prove control with real-time audit trails of data lineage — the documented path data takes through a system — and usage. Such systems must demonstrate exactly how AI systems ingest and process data at any given time.
What is at stake?
Australia stands at what the research calls a critical inflection point in its AI journey. Ambition is present. The constraint is execution: the country's ability to scale AI confidently is being limited by gaps in data and AI sovereignty practice.
The stakes are asymmetric. The organisations that close the control gap will build trust with customers, regulators and their boards, according to the analysis. Those that do not will find themselves increasingly exposed in an environment where accountability demands are only growing.
For Australian enterprises, the 70% figure is less a statistic than a deadline. The gap between data sharing and data governance is measurable, comparative and, on the evidence of this research, still widening.
Source: itbrief.com.au
More from Marcus Bennett
Show full bio
Senior reporter covering companies, deals and strategy in artificial intelligence and machine learning at AI In Context.
718 articles
Related articles
- Only 8% of Australian Firms Have Data Ready for AI at Scale
- BlackFog survey: 47% report sensitive data exposure via AI
- Only 15% of APAC boards feel equipped to oversee major risks
- Two-thirds of Australian firms hit by AI-linked cyber threats: ESET
- 65% of Australian Firms See AI Returns — But Readiness Is Falling