Singapore leads the world in AI maturity, but its workforce also has the highest trust barrier in APAC — a combination that new research from Notion suggests reflects a demanding workforce rather than resistance to AI.

Notion’s Global AI Transformation Study 2026, which surveyed more than 6,000 business leaders and workers across 10 global markets, found that 21 per cent of Singapore organisations are operating at Level 3 or 4 of Notion’s AI Transformation Model, nearly double the global average of 12 per cent and ahead of every other market surveyed.

Singapore closes the leader-worker confidence gap

Globally, the study found a wide disconnect in AI confidence: 49 per cent of leaders report high confidence in their organisation’s AI capabilities, compared with just 23 per cent of workers. Singapore is the exception, with the narrowest gap of any market — 48 per cent of decision-makers and 39 per cent of knowledge workers reporting high confidence.

At the same time, 82 per cent of Singapore knowledge workers say they would use AI more if they trusted it not to make mistakes on important work, the highest trust barrier recorded in APAC. Notion frames this not as reluctance, but as a sign of a workforce that understands AI’s limitations well enough to be selective about it.

“Singapore’s workers are efficiency-maximisers by nature, often their own harshest critics,” said Katsu Nishi, General Manager, APAC, at Notion. “The 82% trust barrier is a signal that Singapore’s workforce expects more from AI than what most organisations are currently delivering.”

Government policy has played a significant role in that lead: 38 per cent of Singapore’s most AI-mature organisations cite it as a major driver of their strategy, nearly five times the rate cited in Japan and the highest of any market in the study.

Adoption remains shallow beyond the leading tier

Singapore’s lead at the top does not yet extend across the broader economy. Ministry of Manpower data cited in the report shows 71.5 per cent of Singapore organisations have not yet adopted AI, and only 3.8 per cent have integrated it into core business processes. High implementation costs (44.9 per cent) and a lack of in-house expertise (42.4 per cent) were the most commonly cited barriers.

Across the global study, three practices most clearly separated the most AI-mature organisations from the rest: defined metrics to track AI’s business impact, governance and oversight for AI outputs, and established guidelines for AI use — each showing gains of 14 to 17 percentage points in APAC.

“The next challenge is making transformation accessible beyond the enterprise tier,” Nishi said. “That doesn’t happen by adding more software licenses. It happens when organisations build the architecture that connects agents, people, and knowledge in one place, with the governance structures that give every level of the business confidence in what AI is actually doing.”

Notion said Singapore enterprise customers including Supabase and OpenGov Products are already deploying its Custom Agents feature, launched in February 2026, across triage, reporting and knowledge management workflows, with APAC now accounting for nearly 30 per cent of all global custom agent actions.