New report warns of widening gap between AI adoption and meaningful organisational transformation
Most businesses across the Asia-Pacific region have yet to translate AI adoption into tangible business outcomes, with the majority still stuck in experimentation or pilot phases, according to a new report.
Findings from the Asia Tech Alliance's (ATA) new report warned that despite APAC leading the world in AI adoption and literacy, a significant gap has opened between uptake and meaningful organisational change.
The report draws on findings from several major surveys. The OECD's 2025 survey of more than 5,000 SMEs found that most businesses deploying generative AI are using it for "non-core activities — simple, one-off, low-stakes tasks — rather than the revenue-producing work that materially shifts business performance."
McKinsey's 2025 State of AI survey found that while a large majority of organisations report regular AI use, "only a small fraction have fully scaled AI across the enterprise; most remain in experimentation or pilot phases."
BCG's APAC data, also cited in the report, warned of firms "plateauing at high adoption and low transformation."
The ATA described this as the "adoption-impact gap," noting that two firms with identical adoption numbers can produce vastly different outcomes.
"One redesigned its workflows and is pulling away competitively; the other is paying for tools nobody meaningfully uses," the report said.
The cost of misapplied AI
The gap has measurable consequences. A Harvard Business School field study cited in the report found that AI boosted productivity when applied to the right tasks, but workers who used it beyond its reliable range were 19% less likely to produce correct solutions than those who used no AI at all.
The report focuses heavily on small and medium-sized enterprises, which account for approximately 97% of enterprises and 85% of employment across Southeast Asia.
It identifies three structural gaps preventing SMEs from moving beyond adoption, including a knowledge gap, where firms adopt AI tools without understanding how they fit into broader business operations.
It also pointed out a capability gap, where owners and managers lack the organisational bandwidth to redesign workflows even after completing AI training, as well as a readiness gap, covering the longer-term challenge of building literacy, fluency, and mastery within leaner teams and tighter resource constraints.
Supply-side training not enough
The report argues that training programmes alone are insufficient.
Gog Soon Joo, Senior Director at the Institute for Adult Learning Singapore, is quoted as warning that workforce policy "often focuses too heavily on supply-side training."
"The demand side also matters. Companies must know how to utilise AI-enabled workers, and workplaces must create roles that leverage AI skills," the senior director added.
"Otherwise, workers may develop capabilities that firms do not know how to use effectively — creating inefficiencies in the labour market."
A new measure of success
The ATA's answer to this challenge is the AI Transformation Compact.
The report states that, rather than another training programme or adoption incentive, the model brings together governments, industry bodies, AI vendors, and employers in a shared, sector-specific commitment to redesign how work actually gets done — with a 12 to 24–month window to show results.
The report states that success should be measured not by licences sold or courses completed, but by whether the work itself looks different twelve months on.
It also recommends tiering publicly funded AI training across foundational literacy, applied fluency, and domain mastery, and calls for an APAC longitudinal panel to track how firms and workers change over time, arguing that cross-sectional surveys cannot adequately capture the distinction between adoption and transformation.
