Businesses in Australia are finding it difficult to scale artificial intelligence (AI) efforts, even though returns on investment are showing promise. While early adopters are achieving up to 29% ROI within two years, most organisations are still running isolated projects without a clear, enterprise-wide strategy. This leads to fragmentation, limiting improvements in productivity.
The recently released National AI Capability Plan has brought renewed focus to a long-standing issue in the Australian economy. While organisations recognise the benefits of AI, few are applying it at scale. Even though AI offers the potential to improve efficiency, fewer than 10% of organisations are using it across their operations. This shows a gap between ambition and actual execution.
Findings from the SAP Value of AI Report, conducted by Oxford Economics, show that 73% of businesses expect to see positive returns from AI over the next three years. However, adoption is still being slowed by factors beyond technology, including unclear processes, slow decision-making and inconsistent ownership. Even with increased efforts in training and upskilling, structural issues continue to prevent full integration of AI into daily operations.
Nearly 70% of businesses say staff are using unauthorised AI tools. This points to growing frustration with internal systems that lack the agility and intelligence employees expect. The rise of so-called "shadow AI" reflects a strong appetite for AI, but without the proper channels in place to support it.
The next evolution of AI, known as agentic AI, which can perform tasks autonomously, is expected to deliver a further 10% ROI boost within two years. Yet only 6% of businesses feel ready to adopt this technology. Without a strong strategy, the introduction of new tools may increase internal fragmentation rather than drive coordinated performance.
Some leading organisations are responding by taking a dual-speed approach. They are using embedded AI within existing systems for widespread, immediate benefits, while also deploying customised solutions in specific areas facing challenges. This approach helps balance broad reach with targeted impact, allowing AI to provide ongoing value over time.
Experts maintain that AI must be a fundamental part of the business strategy, not just the remit of IT departments. Success depends on having accessible data systems, flexible risk management plans and direct involvement from company boards. Without cohesive alignment across these areas, organisations risk slow progress, while more prepared competitors advance from pilot phase to sustained economic impact.
