Singapore businesses expect returns from artificial intelligence to nearly double within two years, even as most admit they are not ready to govern the technology at scale, according to a new study by SAP and Oxford Economics.

The SAP Value of AI Report 2026, which surveyed 2,600 business leaders across 13 countries including 200 in Singapore, found that local firms expect overall AI return on investment (ROI) to rise from 19 per cent this year to 36 per cent within two years. Agentic AI alone is expected to generate US$9.8 million in ROI in Singapore over the same period, almost double last year’s estimate of US$4.9 million.

Confidence in agentic AI outpaces readiness

Eighty-nine per cent of Singapore businesses believe agentic AI could have a moderate to very high impact on their organisation, but only 2 per cent say they are fully prepared for it. AI already supports 27 per cent of tasks in the average Singapore business, a figure expected to reach 47 per cent within two years.

“Singapore businesses are moving from AI experimentation to AI execution, and the expected returns from agentic AI show just how much confidence there is in the technology. But value will not come from AI adoption alone. It will come from connecting AI to business data, processes and governance,” said Eileen Chua, Managing Director, SAP Singapore.

Data, skills and governance gaps widen

The report points to foundational weaknesses that could limit how much of that expected value is actually realised. Data readiness has slipped from 62 per cent in 2025 to 55 per cent this year, and 82 per cent of businesses report challenges with incomplete data — issues already causing rework, delays or backlogs for 81 per cent of firms surveyed.

• Less than half of Singapore companies (45 per cent) have a dedicated AI leader, and only 32 per cent have leadership KPIs tied to AI.
• 79 per cent say internal upskilling is not keeping pace with AI tools.
• Only 10 per cent say their governance processes and frameworks are fully ready to govern AI effectively; just 12 per cent say the same of their skills.
• 27 per cent of companies lack a human-in-the-loop process for agentic workflows, and 30 per cent have no permission or access controls for AI agents.

Two-thirds of Singapore businesses (66 per cent) either agree or are unsure whether they are deploying AI agents faster than they can govern them, a gap SAP warns could compound as autonomous AI use cases spread across the enterprise.

“Agentic AI raises the stakes for enterprise readiness. When AI systems can act across workflows, businesses need clear visibility into where agents are operating, what data they can access, and where human oversight is required. Without that foundation, organizations risk creating more activity without achieving better outcomes,” Chua said.

SAP said the findings point to a widening gap between businesses’ confidence in AI’s financial potential and their operational ability to deploy and govern increasingly autonomous systems at scale.