Organisations in Singapore are moving from AI trials to wider deployment, but governance and control frameworks are lagging behind, according to Insight.
Its research found that 37% of organisations in Singapore are scaling AI across functions and 14% have fully embedded it into operations. The findings are based on a survey of 220 business decision-makers in Singapore and 318 in Australia from organisations with at least 100 employees.
Singapore appears to be ahead of Australia in both adoption and reported returns from AI projects. Some 55% of leaders in Singapore said they were seeing moderate to strong return on investment from AI initiatives, compared with 41% in Australia.
At the same time, the study points to growing unease over how AI systems are managed as they take on more responsibility inside businesses. Insight described this as an "autonomy paradox", in which companies give AI systems more decision-making freedom while oversight structures are still being established.
The tension is also reflected in preparedness levels. Nearly 40% of leaders in Singapore said they were moderately prepared for autonomous AI deployment, while 20% said they were highly prepared.
Governance gap
Mike Morgan, Senior Vice President and Managing Director, APAC, at Insight, said the findings reflected a broader shift in how companies are approaching the technology.
"What we're seeing is an autonomy paradox emerging across organisations in Singapore. Businesses are advancing AI decision-making while simultaneously strengthening the governance, trust and control structures needed to support it," Morgan said.
He added that oversight needs to be built into AI programmes from the beginning rather than added later.
"Having partnered with clients across their AI journey, we see first-hand that governance cannot follow implementation, it must be built from the outset, with continuous oversight embedded from the start," Morgan said.
The research suggests many companies are moving ahead with deployment before all governance questions are settled. That marks a shift from earlier stages of AI adoption, when organisations were more likely to keep projects in pilot or experimental phases.
Leaders are increasingly delegating decision-making to AI systems as they seek to expand use across business functions. The report argues that this is forcing management teams to weigh the commercial case for wider AI use against operational risk and accountability.
Scaling barriers
Technical and financial constraints remain significant obstacles to broader rollout. The biggest barrier identified by respondents in Singapore was integration with legacy systems, cited by 33%.
Cost and infrastructure constraints followed, with 25% saying those issues were slowing expansion. The findings indicate that even where companies are more advanced in adoption, older technology environments still complicate broader AI deployment.
Data readiness was described as stronger in Singapore than in other markets covered by the study, but many organisations are still working on the data foundations needed to support AI at scale. Skills shortages also remain part of the picture as companies try to move from isolated use cases to broader operational deployment.
Trust in AI is also rising, though decision-makers' confidence appears conditional rather than absolute. Half of leaders surveyed in Singapore said their confidence in AI falls in high-stakes scenarios, suggesting that acceptance weakens when the consequences of error are greater.
That distinction may become more important as companies place AI in areas involving core business decisions, customer interactions or sensitive internal processes. It also helps explain why governance, accountability and human oversight remain central concerns even in organisations that have moved beyond early experimentation.
Morgan said businesses were no longer waiting for perfect readiness before handing more work to AI systems.
"Organisations are no longer waiting to be fully ready to delegate to AI; they are already doing so to scale outcomes and stay competitive," Morgan said.
He said the rise in autonomy was increasing pressure on companies to keep control structures in step with deployment.
"Autonomy is rising, but so is the need for strong governance, accountability and trust," Morgan said.
