Organisations miss 55% productivity gains from AI adoption

AI used widely, yet productivity gains remain untapped.

Organisations miss out on about 55% of potential productivity gains due to disconnect between artificial intelligence (AI) adoption and workforce readiness, according to findings from the EY 2025 Work Reimagined Survey.

The survey covered 15,000 employees and 1,500 employers across 29 countries, including a sample of 200 employees and 20 employers from Singapore.

Data showed that nine out of 10 employees use AI at work, where 31% of them use it daily for basic searches, emails, and document summary, whilst only 7% use it to advance their workflow.

Workplace AI anxieties are also contributing to the value gap despite widespread AI adoption. About 50% of Singaporean respondents worry about over-reliance on AI would dull their skills and expertise, whilst 40% are concerned about misinformation driven by deceptive but realistic content at scale.

Most of the respondents also perceived a growth in workloads over the past year, with only 5% globally acquiring sufficient AI training to maximise productivity.

Shadow AI is still prevalent in the workforce despite employers offering internal tools, with 26% of Singapore employees from various sectors still bringing their own AI solutions to work.

In terms of integration of AI in the workforce, only 40% of organisations are able to achieve this.

Samir Bedi, EY Asean People Consulting Leader, said that whilst the city is embracing AI in the workforce, real productivity gains will only come when organisations close the anxiety and skills gap.

He also urged employers to invest in purposeful, hands-on AI training and create a safe environment for experimentation, noting that secure and effective AI tools can reduce the disconnect.

For Singapore’s talent health, there is an eight-point rise from 45 to 53 of out 100, which is an 18% increase that shows greater satisfaction with rewards, development and culture.

Resignation rates in the region dropped to 34%, the lowest in four years, after peaking at 51% in 2021 during the “Great Resignation.”

Leadership is essential for maintaining talent health. Leaders who trust, support, and empower their teams create a positive culture that encourages continuous learning and better AI adoption.

The survey also revealed that employees with over 81 hours of annual AI training report an average productivity gain of 15 hours per week.

However, 55% of them are more likely to leave due to high demand in AI talent and external opportunities outweighing internal promotion cycles.

The report also highlights tensions between human factors and AI integration in five key areas, including AI adoption excellence, learning, talent health, organisational culture, and reward structures. Organisations excelling in these areas achieve a “Talent Advantage” and unlock transformational value from AI.

“The survey shows that AI training is paying off in productivity, but it also creates a new retention challenge," Bedi said. "As AI-skilled talent becomes more marketable, organisations must pair upskilling with compelling career paths and total rewards to ensure that the people they invest in can see a future with them."