Channel NewsAsia reported that a Singapore National Employers Federation (SNEF) survey found people confirmed AI’s final decision in 85% of nearly 3,000 human resources use cases, putting human review at the center of most reported workflows.
The survey covered 1,000 employers in June and July 2026, CNA said. In the other 15% of use cases, AI reached the decision without a person’s sign-off.
In nearly all cases without sign-off, employers reportedly assigned someone who could intervene and change or reverse an AI decision if needed. The survey describes respondents’ reported workflows; it does not by itself establish a legal duty to obtain human sign-off.
CNA said routine or high-volume administrative tasks accounted for 73% of reported cases. Examples included résumé screening and skills assessments.
The reported HR AI use rate was 81% among organizations with more than 200 employees and 65% among small and midsize enterprises with 200 or fewer employees, according to CNA. The survey found more use earlier in the employee lifecycle, including recruitment, learning and workforce planning, and the least use in separation and exit matters.
CNA also reported that SNEF said employers generally avoided AI in cases with more significant consequences for employees. SNEF President Kuah Boon Wee said human oversight should reflect the potential consequences, saying it should “match the stakes.”
A separate measure from Singapore’s Ministry of Manpower (MOM) covered companywide AI adoption. In April 2026, MOM reported that 28.5% of surveyed firms had started adopting AI and 71.5% had not. Fieldwork ran from Jan. 19 to March 31 and covered 2,560 private-sector establishments with at least 10 employees.
MOM counted firms planning AI use, piloting or testing it, integrating it into processes, or having employees actively use it as adopters. Its surveywide measure therefore covers a broader set of activity than SNEF’s reported HR-use figures. MOM also found adoption rates of 23.9% at firms with fewer than 25 employees and 76.4% at firms with more than 500. The surveys use different scopes and definitions, so the government figures are not a direct comparison with SNEF’s HR findings.
