Nearly four in five Singapore IT leaders say inadequate real-time data infrastructure is preventing their organisations from scaling AI initiatives, according to Confluent’s 2026 Data Streaming Report.
The study surveyed 4,625 IT leaders across 14 countries, including Singapore, and examined the challenges enterprises face as they move AI projects beyond experimentation.
Although 75% of Singapore organisations are already deploying or piloting agentic AI solutions, many are struggling with the infrastructure, governance, and organisational capabilities required to support them.
AI projects stall over infrastructure gaps
According to the report, 78% of Singapore IT leaders have encountered at least three challenges when scaling AI. These include insufficient infrastructure for real-time data processing, cited by 78% of respondents, fragmented data ownership at 73%, and insufficient skills and expertise in managing AI at 73%.
These constraints are also affecting agentic AI projects. Among Singapore respondents, 95% said they experience or anticipate problems with data infrastructure and quality, while the same proportion cited legacy system integration. A further 93% identified large language model reliability as a concern.
As a result, 73% reported stalled agentic AI projects, while half said they had abandoned such work entirely. Across APAC, the corresponding figures were 74% and 53%.
“Businesses across Singapore are rapidly embracing AI, strengthening the country’s position as a global leader in AI governance. But as AI systems become more embedded in business processes, trust cannot come from regulation alone, especially given the different regulatory approaches across APAC,” said Greg Taylor, Senior Vice President for APAC at Confluent.
He added that organisations need confidence in their data to power every output, decision and action, placing the responsibility on business leaders to assess whether their data infrastructure is ready to support AI at scale.
Investment shifts towards data foundations
The report found that access to current business information is becoming a higher priority as organisations move AI projects into production. Among Singapore respondents, 86% identified continuous and up-to-date business visibility as a top business priority, compared with 91% across APAC.
Data sovereignty and provenance are also attracting greater attention. In Singapore, 86% of IT leaders said effective management of data sovereignty is important, while 82% highlighted the importance of data provenance and tracking capabilities. Across APAC, those figures stood at 90% and 86%, respectively.
The report also examined how organisations view data streaming platforms in relation to AI. Among Singapore respondents, 90% said such platforms can help address governance, risk, and compliance issues in agentic AI by enforcing data access and usage policies upstream.
A further 91% said data streaming platforms could help address agentic AI challenges by improving the reliability of large language models and ensuring that data remains complete and up to date. Meanwhile, 92% said the platforms can make data more trustworthy, contextualised, and discoverable.
Investment priorities reflect that shift. Data management and governance ranked highest, cited by 90% of Singapore respondents, followed by data streaming at 86% and AI and machine learning solutions at 85%.
The report found that 65% of organisations had created richer and more responsive customer experiences, while 61% reported greater automation and responsiveness in internal processes.
“Most organisations do not have an AI investment problem, they have a data problem. AI systems depend on fresh, accurate and contextual information, but too many are still being built on fragmented data, batch processes, and infrastructure that was not designed for continuous intelligence,” said Shaun Clowes, Chief Product Officer at Confluent.
Clowes said organisations deploying AI across critical business processes can no longer overlook shortcomings in their data infrastructure. He added that AI models need to be connected to real-time business systems and events so they can operate using current, contextual information. Organisations making the most progress, he said, are investing not only in AI itself, but also in the data foundations needed to support it.
