Moving from pilot to production is a crucial step in capturing AI value. AI experimentation is accelerating and about a third (32%) of business leaders in Singapore have moved 40% or more of their AI pilots into production, higher than the global average of 25%.
This is according to a new report from the Deloitte AI Institute, which is based on a survey of 3,235 director to C-suite-level leaders with direct involvement in their companies’ AI initiatives, including 75 in Singapore. It was done between August and September 2025.
Findings show that in the next three to six months, 54% of respondents, both in Singapore and globally, expect to reach this level of deployment.
Organisations are faced with competing priorities: the need to run their core business with current technology while investing in the innovation required to compete in the future.
“Business leaders in Southeast Asia have strong AI ambitions, and they are already reaping benefits, particularly in terms of productivity gains,” said Chris Lewin, AI & Data Capability leader at Deloitte Asia Pacific.
“As they continue to invest in AI, they will need a clear roadmap to guide their transformation beyond incremental optimisation, with effective governance firmly embedded at all levels,” said Lewin. “The full value of the technology will come from reimagining what is possible to achieve strategic differentiation and enduring competitive advantage.”
Beyond productivity
Business leaders in Singapore are seeing benefits from their AI efforts, with 73% reporting improved efficiency and productivity, above the global average of 66%.
Further, 53% of Singapore respondents say that AI has enhanced their decision-making and provided data-driven insights – similar to the global average of 53%.
However, the report notes that more can be done in terms of AI-driven transformation. Only 33% of leaders in Singapore and 30% globally say their organisations are redesigning key processes around AI while keeping their overall business model intact.
Even fewer still (28% in Singapore) use AI to fundamentally reinvent their core processes and business models, although this is higher than the global average (17%).
In terms of barriers faced when integrating AI into roles and workflows, the top challenges cited by Singapore leaders are regulations and compliance (27%), AI skills and knowledge (24%), followed by high implementation costs and insufficient technology infrastructure (both 15%).
To prepare their workforce for the impact of AI, businesses in Singapore are focused on building AI fluency as their top priority (53%), similar to the global average (53%).
There is a greater level of awareness among Singapore respondents about the need for broader job redesign, as 47% of them are redesigning career paths and mobility strategies due to AI adoption, compared to 33% globally.
Governance and growth
Agentic AI is poised for rapid growth with close to three-quarters of companies (72% in Singapore, 74% globally) planning to deploy agentic AI in several operational areas within two years, significantly higher than today (14% in Singapore, 23% globally).
The use cases are broad, as respondents identify customer and support services (24%), supply chain and logistics management (15%), and marketing and sales (15%) as the top three areas where agentic AI will have the greatest impact on their industry.
Despite their enthusiasm for agentic AI, only 14% of leaders in Singapore report having a mature model for agentic AI governance, below the global average of 21%.
Half (50%) of respondents in Singapore use a mix of public and internal proprietary frameworks to assess the risk and performance of autonomous agents.
Deloitte’s report notes that, as AI agents take actions directly – compared to conventional AI systems that provide recommendations for humans to act upon – agentic AI requires new approaches to governance.
These include clear boundaries for agent autonomy, real-time monitoring systems to track agent behaviour, as well as audit trails to capture the full chain of agent actions.
The rise of physical AI
Physical AI, defined as AI systems that perceive the real world, make decisions, and drive physical actions through machines or control systems, is increasingly being embedded in business operations.
In Singapore, 84% of businesses expect to use physical AI in their operations within the next two years, up from 53% today, suggesting a new wave of industrial automation.
Business leaders see value in a variety of physical AI applications. The areas that they believe physical AI will have the greatest impact on their industries include digital twins (25%), collaborative robotics (24%) and intelligent security systems and/or smart monitoring (21%).
To help ensure public acceptance and trust of these solutions, Deloitte’s report notes that businesses should focus on making them secure, interoperable and resilient against disruptions and cyber threats.
Sovereign AI
Resilience in the age of AI increasingly depends on sovereign AI readiness, and organisations are taking note. The majority (77%) of businesses in Singapore say that data residency and in-country or in-region compute considerations are moderately to extremely important to their strategic planning.
Along a similar vein, 57% are moderately to extremely concerned about their over-reliance on foreign-owned AI technologies and compute provision.
To prepare for sovereign AI requirements, Deloitte recommends that companies assess which data and workloads must remain within national or regional boundaries, determine where local model hosting is mandatory, and clarify how transparency, auditability, and documentation standards differ across markets.
They should also establish clear policies for data residency, model retraining, and cross border flows, supported by infrastructure capable of meeting complex regulatory requirements.
