While AI adoption has become mainstream, legacy data architectures are increasingly limiting organisations’ ability to scale AI securely, efficiently, and cost-effectively.
This is according to a report based on a survey commissioned by Cloudera and fielded by Wakefield Research. It covered 1,500 enterprise architects, cloud infrastructure leads, and data architects in companies with at least 250 employees.
Respondents were based in Spain, Singapore, South Africa, United States, Canada, Brazil, the United Kingdom, India and Japan. The survey was done June 5-22, 2026.
The findings point to a fundamental shift in enterprise IT architecture. While 77% of organisations are actively using AI, nearly all (95%) have delayed or canceled AI initiatives over the past year because of data governance, compliance, or regulatory challenges.
To overcome these challenges, 72% say their current data architecture requires a significant overhaul to meet future AI requirements. A similar pattern is emerging across Asia Pacific, where 68% say their existing architecture requires significant change, suggesting that today’s infrastructure was not built for the demands of modern AI.
Together, these findings highlight what Cloudera calls “The Great AI Re-Architecture”—the mass transition from legacy data architectures toward hybrid environments that enable organisations to bring trusted AI to trusted data, wherever it resides.
“Many enterprises are discovering that the architectures built for traditional analytics weren’t designed for the scale, governance, and flexibility AI demands today,” said Sergio Gago, CTO at Cloudera.
“Success will depend on building a data foundation that gives organisations the freedom to run AI wherever it makes the most sense, without compromising control or security,” said Gago.
Three-quarters (75%) of respondents say AI integrations have changed their organisation’s data storage and architecture practices, while 84% report increased infrastructure costs driven by AI workloads.
Together, these findings suggest organisations are rethinking not only where data lives, but how it is managed, governed, and delivered to AI systems.
Nearly three-quarters (73%) of respondents say AI has made data governance more complex, with the challenge also evident across Asia Pacific at 61%. These pressures are already affecting deployment: more than half (55%) of the organisations across the world report delaying or canceling more than six AI projects over the past 12 months due to governance, compliance, or regulatory challenges, while 92% of APAC organisations have delayed or canceled at least one project.
In the region, data security, governance and compliance requirements are also the leading driver of AI infrastructure change, cited by 46% of respondents.
The challenge is compounded by increasingly distributed data. Nearly every respondent (97%) reports moving data between environments at least monthly, making consistent governance across cloud, private cloud, on-premises, and edge environments essential for scaling AI securely.
Organisations are increasingly adopting hybrid architectures to balance performance, governance, cost, and flexibility, all of which are critical to modern AI success.
Two-thirds (66%) of respondents globally and 64% across APAC say they have moved AI workloads from public cloud environments back to private cloud or on-premises infrastructure during the past year, signaling a broader shift toward hybrid architectures that allow organisations to run AI workloads where they perform best.
Looking ahead, organisations are investing across cloud, on-premises, edge, and hybrid environments rather than relying on a single deployment model.
One-quarter (25%) say they plan to prioritise a hybrid-first architecture over the next two years, reinforcing that the future of enterprise AI will be defined by flexibility rather than a single infrastructure strategy.
Enterprises in Singapore are focusing on high-performance AI environments, matching their infrastructure strategies with sophisticated AI adoption and corporate innovation.
Highlighting a dynamic review of optimal runtime environments, 43% of Singaporean organisations have transitioned AI workloads out of the public cloud, and an additional 36% are evaluating this shift. Investment patterns also highlight edge infrastructure as a primary target, with 35% of local respondents projecting higher edge spending over the coming two years.
Concurrently, balancing processing capabilities with operational oversight remains crucial as AI initiatives expand, with 41% identifying data security, governance, and compliance as the primary catalysts for modifying AI infrastructure.
“Across the Asia Pacific region, the conversation is transitioning from merely selecting an AI host to determining which environment yields the superior business results for specific workloads,” said Remus Lim, SVP in Asia Pacific & Japan at Cloudera.
“Singapore reflects this evolution as enterprises collectively evaluate latency, performance, expenditure, and governance when defining their AI deployment strategies,” said Lim. “To scale AI breakthroughs effectively, businesses must possess the agility to transition data and AI seamlessly among cloud, edge, and on-premises setups while maintaining strict governance and operational command.”
