Australian organisations are pouring investment into AI. Boards are approving budgets, IT teams are piloting new models, and business units are racing to embed AI into everyday operations. Yet new IDC research points to an uncomfortable truth: the technology isn't the bottleneck. The data underneath it is.
.png&h=420&w=748&c=0&s=0)
Across industries, the gap between organisations generating real business outcomes from AI and those stuck cycling through pilots has little to do with which large language model they've chosen. It has everything to do with whether their data, governance, and infrastructure foundations can actually support AI at scale.
What new IDC research reveals about AI readiness
IDC's findings draw a sharp line between two groups: IT Modernisation Experts and Explorers. The difference isn't ambition or budget. It's maturity. Experts have done the unglamorous work of unifying data sources, establishing governance, and modernising the systems that hold their most valuable information. Explorers, by contrast, are often still experimenting, layering AI tools onto infrastructure that was never designed to support them.
The business outcomes diverge sharply as a result. According to IDC, Modernisation Experts extract 59% of the potential business value from their data, compared to just 22% for Explorers. That's not a marginal advantage; it's the difference between AI as a genuine competitive asset and AI as an expensive experiment that never quite delivers.
For IT leaders at organisations running mainframe-centric environments, this finding should land as validation rather than surprise. We've long argued that successful modernisation isn't about ripping and replacing core systems. It's about making the data and logic within them accessible, governed, and ready for what comes next.
The data accessibility gap identified by IDC
If data maturity determines AI outcomes, accessibility is where that maturity gets tested. IDC research shows Modernisation Experts increased the internal data accessible to AI by 27%, while Explorers managed just 14%. That gap compounds over time. Every quarter an organisation delays unifying its structured and unstructured data, AI initiatives built on incomplete information move further from delivering reliable results.
This is particularly relevant for enterprises operating complex IBM Z® environments. Decades of embedded business logic and transactional data represent enormous value, but only if that data can be surfaced, cleaned, and connected to the tools teams are using to drive decisions. Organisations that treat data accessibility as a one-time project rather than an ongoing capability will keep falling into the Explorer category, regardless of which AI platform they adopt next.
Why new IDC research challenges conventional thinking about legacy systems
For years, the conventional narrative framed mainframes and legacy platforms as liabilities to be modernised away from. IDC's research challenges that thinking directly. The real opportunity isn't retiring these systems; it's evolving them from systems of record into systems of intelligence.
A system of record simply stores transactions. A system of intelligence makes that data available, governed, and ready to feed AI-driven decision-making in real time. Replatforming and refactoring core applications, rather than rewriting them from scratch, lets organisations preserve decades of proven business logic while making it accessible to modern analytics and AI tooling. This incremental, capability-led approach reduces risk and avoids the disruption that often accompanies large-scale system rewrites, while still unlocking the data value AI depends on.
IDC research shows governance is becoming an AI imperative
Data maturity without governance is its own kind of risk. IDC's research reinforces that strong governance, security, and access controls aren't compliance checkboxes. They're prerequisites for scaling AI responsibly. Without them, organisations expose themselves to exactly the kind of ungoverned technology use that erodes trust, invites regulatory scrutiny, and produces inconsistent, unreliable AI outputs.
For Australian enterprises in regulated sectors, this is especially pressing. Robust data governance determines whether AI-driven insights can be trusted, audited, and defended. It's the foundation that allows IT leaders to move from cautious experimentation to confident deployment.
The agentic AI readiness challenge
Agentic AI, systems capable of taking autonomous action rather than simply generating recommendations, raises the stakes considerably. IDC's research suggests agentic AI will amplify existing modernisation maturity gaps rather than close them. Organisations with weak data foundations won't simply see slower AI adoption; they'll see agentic systems acting on incomplete or poorly governed data, with consequences that compound across the business.
This is the readiness challenge IT leaders need to confront now, before agentic AI moves from pilot to production. The organisations best positioned to benefit will be those that have already modernised their data estates and built governance into the foundation, not bolted it on afterward.
Build the foundation, not just the pilot
The question Australian IT leaders should be asking isn't "Are we adopting AI?" Nearly everyone is. The question that actually separates Experts from Explorers is: Have we built the foundation that allows AI to succeed?
Rocket Software has spent decades helping organisations in regulated, mainframe-centric industries modernise data, strengthen governance, and extend their core systems into the cloud without disruption. As IDC's research makes clear, that foundational work isn't a prerequisite to postpone. It's the differentiator that determines whether AI investment turns into genuine business value.
Read the full IDC research to explore the complete findings on modernisation maturity and AI readiness. And if you're ready to see how Rocket Software can help your organisation build the foundation AI demands, learn more at rocketsoftware.com
Assess Your Organization’s IT Modernisation Maturity | Rocket Software
