AI Usage Alone Fails to Measure Productivity, New Study Finds

What happened

A new study reported by TechRadar finds that people who use AI at work are putting in up to 90 minutes of extra time every day, despite the productivity gains that AI tools have been promoted as delivering.

The coverage frames the findings as a challenge to adoption as a measure of success. Its central argument is that the goal cannot simply be AI usage. Counting how many people use a tool says little about whether it is saving time, reshaping workloads or extending the working day.

The reporting presents the study as evidence of a gap between the efficiency case made for AI and the day-to-day experience of the people using it. The source material does not provide the sample, methodology or timeframe, so readers should treat the 90-minute figure as the study's headline finding rather than a fully detailed benchmark.

Why it matters

For leaders in experience, AI and digital transformation, the finding suggests that deploying a tool and realising its benefit are separate problems. An AI tool may remove some steps while adding others, such as reviewing outputs, correcting errors, switching between systems or handling new expectations for speed. When those costs are not visible, organisations can report high adoption while employees quietly absorb the difference in their own time.

The implications extend beyond the employee. Staff who are working longer than planned are the same people shaping service quality, handling complaints and making judgement calls that customers feel. Operators that treat AI as a pure efficiency play risk trading short-term output for fatigue, slower responses and inconsistent service.

The Renascence take

Most organisations are measuring the wrong thing. Adoption dashboards reward activity, not outcomes, and they make it easy to mistake busyness for value.

Stop asking how many people use AI and start asking how much time it returns and where that time goes. Treat the hidden effort of checking, correcting and context-switching as a design flaw, not a training gap. Measure the working day before and after deployment, including out-of-hours work, and tie any AI rollout to a clear reduction in effort for the people serving customers. If the tool adds work, redesign the workflow before you scale it.