Organizations across Asia-Pacific (APAC) are confident about their artificial intelligence (AI) progress, yet returns have been limited, as investments in the technology remain focused on small improvements rather than transforming the way they operate, according to a study by professional services company Avanade.

The findings come from Avanade's “Build Beyond Limits: AI Reality Report,” which surveyed 3,600 business leaders from various sectors across 11 markets, including North America and Asia-Pacific and Japan. The study pointed to a growing gap between organizations' confidence in their AI maturity and their actual progress.

“The global survey found that 77% of organizations describe themselves as Established, Advanced or AI-first in their AI journey, while 72% have high confidence their teams are using AI effectively,” Avanade said. “And yet, only 18% say they have fundamentally reshaped how their teams operate around AI, and 75% believe organizations publicly present themselves as more AI mature than they really are.”

The same pattern held in Asia-Pacific, where 76% of the region's 1,400 respondents — nearly four in ten of the global sample — described their teams as Established, Advanced, or AI-first in their AI journey, and 73% said they have high or complete confidence that they are using AI effectively to deliver their strategy and goals. Only 17%, however, said they have fundamentally reshaped how they operate to leverage AI, leaving a 56-percentage-point spread between confidence in AI use and evidence of fundamental operational change.

According to the study, 79% of APAC business leaders said too much of their organization's AI investment is directed toward incremental improvements rather than transformational opportunities.

“Incremental AI improvements can create real value, but improving existing operations and fundamentally transforming a business are different ambitions. APAC leaders need to be clear about where each applies, with investment decisions and measures of success to match. Where transformation is the goal, that means redesigning workflows, responsibilities and decision-making, with clear ownership of outcomes. The real test is whether the intended change happens and its value can be demonstrated,” said Bhavya Kapoor, president of Avanade Asia Pacific.

Some 76% of respondents also said their own organization publicly portrays itself as more AI mature than it is, the study showed, closely tracking the global figure of 75%.

“Separately, only 43% of APAC respondents say their organizations use business-value and RoI (return on investment) targets to judge whether they are making meaningful progress with AI,” the company said. “This suggests that many organizations may still be evaluating AI activity and adoption without consistently linking it to business outcomes and value creation.”

It added that 68% of respondents said their organization has yet to formally calculate the financial return on its AI investment.

Avanade said business leaders need to be able to distinguish between AI investments that only improve existing operations and those that can fundamentally change value creation, and should set clear objectives, accountability, and measures of success.

“Too many organizations are measuring AI progress by activity rather than impact. Deploying tools, running pilots and improving productivity are important milestones, but they are not the destination. What remains to be seen is whether organizations are truly changing and investing in how decisions are made, how work gets done and how value is created through AI. AI-native challengers are building these capabilities from day one, and established organizations must think just as boldly about the future they want to create,” said Ben Beath, chief AI officer at Avanade.

Against that backdrop, the report leaves APAC organizations with two observable markers going forward: whether AI investment shifts from incremental improvement toward redesigned workflows and decision-making, and whether business-value and RoI targets become a consistent yardstick for measuring AI progress.

— Bettina V. Roc, BusinessWorld. Original report: Asia-Pacific organizations say AI transformation still limited