Singapore CEOs face AI accountability test as control gaps persist
Dataiku’s survey finds Singapore CEOs under pressure to prove AI returns while governance and explainability gaps slow adoption.
Singapore CEOs are under growing pressure to show that AI can deliver measurable business returns, even as many remain cautious about giving the technology more control over business decisions.
Table Of Content
A Dataiku study conducted by Harris Poll found that 86% of Singapore CEOs say their role is at risk if their company fails to deliver measurable gains from AI by the end of 2026. The Global AI Confessions Report: CEO Edition 2026 surveyed 900 CEOs worldwide, including 100 in Singapore.
The findings point to a practical tension inside enterprises. Dataiku found that 89% of Singapore CEOs would stake their job on the success of AI initiatives, yet 33% would still not allow AI to make decisions without human approval. AI is becoming tied to leadership performance, but trust in how it acts remains limited.
Accountability is pushing CEOs deeper into AI decisions
The pressure around AI is becoming personal for senior leaders. According to the survey, 83% of Singapore CEOs believe a fellow CEO will be ousted because of a failed AI strategy or an AI-driven crisis.
That does not mean AI has displaced every other business priority. In Singapore, 62% of CEOs rank AI strategy as a top or high priority, but only 13% say it is their top priority. The figures suggest that AI is firmly on the leadership agenda, while still competing with other commercial and operational demands.
As accountability rises, CEOs appear to be getting more involved in decisions that were previously more likely to sit with technology teams. Dataiku found that 78% of Singapore CEOs say their involvement in AI-related decisions has increased, while 59% participate in most AI-related decisions.
That involvement is also creating friction over technology choices. Some 64% of Singapore CEOs said they challenged AI vendor or platform decisions made by their CIO or others in the past year. For companies deploying AI across teams and workflows, vendor selection is becoming a leadership issue because the consequences of failure now sit closer to the chief executive.
Governance gaps are slowing AI deployment
The survey also shows why CEOs are reluctant to hand over more decision-making authority to AI systems.
Dataiku found that 95% of Singapore CEOs believe at least some employees are using generative AI tools without approval. That creates a governance problem for companies trying to manage how AI is used, what data enters these systems, and whether outputs can be traced back to approved tools and processes.
Legal and reputational concerns add to that caution. Some 78% of Singapore CEOs worry that AI agents could create legal exposure, while 59% believe poor explainability could trigger a crisis that damages customer trust or brand credibility.
Those concerns are already affecting production plans. While 81% of Singapore CEOs are confident they will be able to deploy AI agents in full production in 2026, 88% of CIOs in Singapore said gaps in traceability or explainability have already delayed or stopped AI projects from reaching production.
That gap between executive confidence and implementation friction is the clearest thread in the findings. CEOs are expected to prove AI’s value, but production use still depends on whether organisations can explain decisions, trace outputs and set clear approval lines.
Board visibility raises the stakes
AI decisions are also becoming more visible at board level. Dataiku found that 94% of Singapore CEOs are comfortable disclosing AI-driven decisions to their boards, in line with the global average.
That visibility can help normalise AI use in strategic decisions, but it also raises the standard for governance. Once AI-supported decisions are discussed at board level, companies need clearer evidence of how those decisions were reached, who approved them and whether the systems involved were properly controlled.
The wider survey suggests this tension is not confined to Singapore. Dataiku said global confidence in deploying AI agents at scale fell from 41% to 31%, even as CEOs continue to link AI success with business performance.
For Singapore enterprises, the issue is less about whether AI will be adopted and more about whether it can be governed well enough to support decisions that leaders are prepared to defend.
