AI at Work: Why Productivity Gains Depend on People

AI is helping employees save hours every week. But research suggests that much of this time is being spent reviewing and correcting AI-generated work.

The productivity gains organisations are looking for

According to global research by Workday, 85% of employees say AI saves them between one and seven hours per week. As highlighted in IT Brief's report on the AI productivity paradox, these findings raise an important question: are time savings translating into meaningful improvements in productivity and business outcomes?

The study, Beyond Productivity: Measuring the Real Value of AI, surveyed 3,200 full-time employees at organisations with annual revenues exceeding US$100 million across North America, Asia-Pacific, and EMEA. Conducted by Hanover Research, it examines how employees experience AI and whether its productivity gains deliver meaningful value.

The hidden cost of AI productivity

The same research found that nearly 40% of the time saved through AI is offset by rework, including correcting errors, rewriting content, and verifying AI-generated outputs.

In practical terms, for every 10 hours of productivity gained through AI, approximately four hours are spent reviewing and refining its output. Only 14% of employees consistently report clear, positive net outcomes from their AI use.

Human oversight remains essential. Among daily AI users, 77% review AI-generated work as carefully as work produced by colleagues, or even more carefully.

This highlights an important distinction: faster output does not automatically mean better results. Employees still need the judgement to assess accuracy, identify gaps, and determine whether AI-generated content is fit for purpose.

Why AI adoption is also a workforce skills challenge

Workday's research found that in 89% of organisations surveyed, fewer than half of job roles had been updated to reflect AI capabilities. It also identified a gap in training: while 66% of leaders consider skills training a priority, only 37% of employees experiencing the highest levels of AI-related rework say they have access to it.

Making AI tools available is only part of the equation. Employees also need to know how to provide relevant context, assess generated information, recognise inaccuracies, and apply their own judgement.

Similar challenges are emerging across the region. Our article on AI adoption and workforce capability in APAC explores the growing importance of developing the skills needed to use AI effectively.

Beyond technical proficiency, collaboration and change management are also important to AI readiness, particularly as new technologies reshape how teams work together.

What organisations can learn from better AI outcomes

The research offers insights into how organisations can make better use of the time AI creates.

Among employees who consistently report positive AI outcomes, 57% use the time saved to deepen their analysis and strategic thinking, while 79% have received increased skills training.

Workday also found that organisations allocate their AI-related productivity gains in different ways: 39% reinvest them in technology, 30% in employee development, and 32% absorb them through increased workloads.

These findings highlight an opportunity for businesses to look beyond technology investment. Developing employee capabilities and providing managerial support can help teams integrate AI more effectively into everyday work.

As explored in our article on the role of managers in AI adoption, leadership plays an important role in helping employees adapt to new ways of working.

Turning AI adoption into meaningful outcomes

For organisations introducing AI, the question should extend beyond how much time a tool saves. It should also consider what happens to that time afterwards.

Are employees equipped to evaluate AI-generated outputs? Have workflows evolved to reflect new capabilities? Is saved time being used to improve work quality, strengthen skills, and support more strategic activities?

A skills-first approach begins by understanding existing workforce capabilities, identifying where AI can add value, and providing practical training to help employees use it responsibly and effectively.

It also requires organisations to rethink how work is structured. As discussed in our analysis of AI and operating model redesign, sustainable AI adoption depends on aligning employee capabilities with workflows that allow those skills to deliver value.

Ultimately, AI productivity is not just about completing tasks faster. It is about combining technology with human judgement, continuous learning, and better ways of working to create meaningful and sustainable business outcomes.

The next stage of AI adoption is not just about giving people access to better technology. It is about equipping them with the skills to make the most of it.

‍