A new study by ServiceNow has revealed a growing disconnect between enterprise ambition and execution when it comes to artificial intelligence (AI) in Asia Pacific (APAC).
The study showed Tuesday that while business leaders are optimistic about AI’s potential, most organizations are reducing their AI budgets.
This is a sign that strategic misalignment, governance gaps, and fragmented deployment are stalling returns and undermining confidence.
According to the Enterprise AI Maturity Index, organizations in Singapore (-4 percent), and Japan (-3.3 percent), Australia (-3 percent), and India (-2.1 percent) all reported year-over-year declines in AI spending as a percentage of their overall technology budgets.
These cutbacks come even as C-suites express growing optimism about AI’s potential.
While interest in AI remains high, the lack of alignment and strategic clarity is becoming a major roadblock.
Only an average of 39 percent of enterprises across four markets in APAC say they are operating with a clear, shared AI vision for business transformation.
India leads with 52 percent, but Singapore (34 percent), Australia (33 percent), and Hong Kong (30 percent) trail behind.
In parallel, visibility into AI deployment remains limited. Only 40 percent of APAC organizations on average say they have strong visibility across functions.
India once again leads (51 percent), while Singapore (36 percent) and Hong Kong (31 percent) fall behind.
“You can’t steer what you can’t see. Enterprises are pushing forward with AI, but without a unified vision or clear line of sight across the business, they’re essentially flying blind,” said CK Tan, APJ Innovation Officer, Singapore at ServiceNow said.
The study also found that as enterprises across APAC race to adopt AI, many are deploying it in fragmented ways that outpace their ability to govern and scale responsibly.
On average, 68 percent of enterprises across the four APAC markets build and deploy AI through multiple task forces.
While this allows experimentation, it introduces risk when governance structures are missing.
Governance has emerged as a critical weak point, the study showed.
More than half of enterprises in Australia (57 percent), Singapore (56 percent), and India (51 percent) have not made significant process putting formal governance frameworks in place to guide AI development and use.
Without a common set of controls, organizations risk duplicating efforts, creating inconsistent experiences, and exposing themselves to data and compliance vulnerabilities.
Employees are also increasingly feeling the impact of rapid AI deployment, with growing anxieties around job security and organizational risk.
In Australia (60 percent) and India (57 percent), a majority of enterprises report that their employees have expressed concerns about job insecurity due to generative AI than almost anywhere else in the world, with Singapore not far behind (54 percent).
At the same time, employees are become more nervous around AI-related issues such as data misuse and model hallucinations, reflecting broader concerns about trust, transparency, and oversight.
“Many enterprises are building isolated AI capabilities without the connective tissue needed to scale responsibly. As deployments grow more complex, governance will become the difference between competitive advantage and operational risk,
“Leading with transparency and putting people at the center of AI is essential to building trust and unlocking long-term value,” added Tan.
The study also showed that organizations that invented entirely new AI-human workflows saw far greater business outcomes than those who simply layered AI on top of existing processes.
In Singapore, companies that reimaged workflows were around three times more likely to see improvements in efficiency and employee experience.
In India, productivity gains were twice as likely for those who adopted new workflows, while in Hong Kong, reimagined workflows led to better outcomes around twice as often across risk management and customer/employee experience.
Despite mounting complexity, most APAC firms are still adding AI on top of existing systems rather than consolidating tools – a practice driving “solution sprawl”.
In Australia and Hong Kong, 61 percent of enterprises say they are layering on new AI solutions rather than replacing them with more integrated options, while at least one in two in both markets are mainly adding single-purpose solutions to boost their AI capabilities.
In Singapore, this number is slightly slower at 50 percent.
Singapore invests boldly in AI but businesses struggle to scale enterprise-wide: IBM study