As enterprises across Asia Pacific rush to embrace artificial intelligence, many are discovering that enthusiasm alone is not enough. While AI projects are multiplying across industries, only a small number of organisations have the data, governance and infrastructure needed to scale them successfully.
In conversation with iTNews Asia, Matthew Hardman, APAC CTO at Hitachi Vantara, explains that while AI adoption is accelerating across Asia Pacific, only a small proportion of enterprises are truly AI-ready. He shares his advice on what organisations must do to move from experimentation to measurable outcomes.
“AI is now a top priority for almost every organisation we speak to, but only a much smaller number of organisations are truly ready to support it at scale,” Hardman said.
According to Hardman, Asia Pacific is no longer in the early experimentation stage of AI. Instead, the region is now moving into a broader acceleration phase, although progress varies widely across countries.
“People often talk about APAC as if it is one big market, but the reality is that the region is incredibly diverse. There are very different levels of maturity and this is reflected in the AI projects organisations are pursuing,” he added.
He pointed to Singapore and Australia as the two most mature markets in the region, where enterprises are moving quickly from pilots to real deployment. China is also advancing rapidly, driven by its own domestic AI ecosystem.
Enterprises want AI, but few are truly ready
Although enthusiasm for AI is high, Hardman believes readiness remains the biggest obstacle. Hitachi Vantara’s research found that between 95 and 97 percent of organisations consider AI a top priority. However, only around one-third believe their infrastructure is mature enough to support it.
The biggest barrier is not lack of ambition, but a lack of foundational readiness. He identified financial services, healthcare, retail and telecommunications as the industries currently leading in AI maturity.
“Financial services have very robust data practices and they understand how to work with models and analytics. That gives them an advantage when it comes to AI,” he added.
Healthcare and retail are also moving quickly, while telecommunications companies are increasingly using AI because of the scale and complexity of their networks.
However, Hardman argued that AI maturity is no longer determined purely by industry.
The organisations that are lagging are not necessarily tied to one vertical. The common factor is that they are still being held back by legacy infrastructure and poor readiness.
- Matthew Hardman, APAC CTO at Hitachi Vantara
Boards are chasing AI, but not always business value
Hardman warned that many organisations are still treating AI as a box-ticking exercise rather than focusing on business outcomes. Instead, enterprises need to define exactly what business problem they are trying to solve, who the customer is, and how AI will improve that experience.
“The real value comes when AI improves customer experience, drives innovation, increases productivity or creates competitive advantage,” Hardman said.
As organisations rethink their AI strategies, he said many are moving away from isolated data silos and towards unified data platforms. He believes consistency across environments will become critical as enterprises increasingly adopt hybrid AI models.
“If you run something in the cloud, it should look and operate similarly to how it works on-premises. That consistency reduces complexity and makes organisations more prepared for future workloads,” Hardman explained.
How to build a safe roadmap for AI
For CIOs and CTOs building an AI roadmap, Hardman said the first step is not choosing a model or a platform. “You have to start with data governance. If your foundation is not strong, everything you build on top of it becomes risky.”
Enterprises now need governance frameworks that extend beyond data to include AI models, usage policies and operational controls. He also urged organisations to establish responsible AI frameworks that clearly define which tools employees can use and why.
Hardman said enterprises looking to move beyond AI pilots should focus on three immediate priorities. First, to build a strong data foundation, identify the business use cases that matter most and ensure governance and security are embedded from the start.
He also pointed to the fourth warning to not ignore the technical debt. “Many organisations are so focused on building new AI capabilities that they forget to retire old systems. If you do not address technical debt, you end up carrying the cost, risk and complexity of legacy infrastructure into the future,” he explained.
