Singapore – Singapore businesses are moving towards wider artificial intelligence adoption, but gaps in infrastructure, data governance and security are creating challenges for organisations seeking to deploy AI at scale, according to a new study from Dell Technologies.
The study surveyed 2,950 business and IT decision-makers across 35 countries, including 100 respondents from Singapore. The findings suggest that while organisations increasingly recognise the potential of AI, many are taking a more cautious approach to new deployments because of concerns around business outcomes, risk and operational readiness.
Nearly six in 10 Singapore organisations (58%) said they cannot keep pace with the speed of technological change, while 60% lack an actionable roadmap covering AI, data and security. At the same time, 93% said macroeconomic pressures have made them more selective and focused on outcomes when making technology investments.
The study found that 79% of respondents would not proceed with new AI initiatives without proven business outcomes and a defined risk management plan.
“Singapore businesses aren’t questioning AI’s potential anymore; they’re drawing a hard line on ROI and risk,” said Andy Sim, Vice President and Managing Director for Singapore at Dell Technologies.
“Nearly eight in ten local enterprises won’t greenlight new AI projects without proven business outcomes and a clear risk plan. Ambition is outpacing readiness, and shiny AI software can’t overcome legacy infrastructure, data security gaps, or rising energy constraints. Until companies resolve these behind-the-scenes bottlenecks, most AI projects will stay stuck in pilot mode,” he continued.
Data governance remains a concern
Data sovereignty and security are also influencing how businesses approach AI deployment. Eighty per cent of Singapore respondents said their data and intellectual property are too valuable to place in third-party generative AI tools, while 86% said regulatory compliance is shaping how and where they use data for AI.
The research found that 79% of organisations have slowed or paused AI adoption because of security and compliance concerns. Nearly three-quarters of respondents also said they require technology partners to demonstrate clear risk management plans before proceeding with new initiatives.
Infrastructure creates another barrier
The study points to storage and data access as a more significant technical constraint than computing capacity for many organisations. Some 82% of respondents identified storage performance and data access as their primary technical bottleneck, while 76% said their data centres are not yet ready to support AI workloads at production scale.
In response, 88% plan to consolidate onto fewer, more capable platforms, while 94% said integrated cyber-recovery capabilities are now a baseline requirement.
Energy availability is also emerging as a consideration in AI infrastructure planning. Eighty-nine per cent of respondents cited power limitations as a core purchasing factor, and 86% said they would only scale AI if energy consumption can be kept under control.
The research also highlights the growing use of AI tools outside centrally managed enterprise systems. Ninety-two per cent of respondents reported that employees use personal devices to access AI tools for work, while 73% said they had identified unapproved AI tools on company devices.
A further 77% said teams are expensing AI tools on departmental credit cards without IT oversight.
At the endpoint level, 87% of organisations said legacy PCs are limiting adoption of AI-enabled applications. Meanwhile, 93% believe AI PCs will play a central role in their future workplace strategies.
Businesses seek more strategic technology partnerships
The findings also indicate a shift in how organisations assess technology providers. Across the Asia-Pacific, Japan and China (APJC) region, 83% of organisations said they are replacing transactional vendors with strategic partners that can help develop multi-year roadmaps and share responsibility for business outcomes.
Ninety per cent of respondents said their existing technology partners should be doing more, while 95% expect security to be embedded into every proposal rather than treated as a separate consideration.
