Few workers have progressed beyond experimentation or task-level assistance when it comes to AI
Most organisations are not yet seeing operational change due to AI adoption, a new report has found, attributing it to the confined usage of the technology at work.
The number of AI users that have successfully embedded AI across multiple workflow steps has grown by 36% to 2,369 users, according to findings from the ActivTrak Productivity Lab.
Despite this growth, the report underscored that they represent just two per cent of the 120,620 workers covered in the report.
The numbers come in the wake of longer AI usage at work, averaging six hours and 34 minutes of productive hours daily in Q2 2026.
But usage has yet to progress beyond experimentation, with 27% of AI users still using the technology to consult it for questions or information, but does not delegate work to it.
Another 14% of AI users also use it just for drafting content, generating ideas, or completing routine tasks.
The findings suggest that AI distribution is widespread, but redesigning work to accommodate the technology is falling behind.
"Most teams have AI access. The real question is whether it's changing how work gets done," said Gabriela Mauch, Chief Customer Officer and Head of the ActivTrak Productivity Lab.
"For most organisations, the answer is not yet. The next step is understanding where work concentrates, where tasks repeat and where AI can genuinely transform workflows."
Advanced users of AI have higher retention and healthier utilisation of the technology compared with individuals in the earlier stages of adoption, according to the report.
The small segment of the workforce embedding AI into their workflows maintains at 69% healthy utilisation, with four to nine daily productive hours.
They also have a shorter average workday span of six hours and 46 minutes, compared with the seven hours and five minutes of users still using AI for task execution.
How to improve AI adoption maturity
ActivTrak has argued that the problem lies in the lack of workflows, enablement, and visibility to progress AI adoption.
It underscored that it is a mistake to treat all employees as if they are all on the same AI adoption stage, and warned against treating more AI tools as a signal of higher adoption maturity.
It advised leaders to identify their team's AI maturity stage in order to implement effective intervention measures.
"Advancing AI adoption maturity is not a technology problem. It isn't solved by deploying more tools, running another training program or adding AI features to existing licenses," the report read.
"It advances when leaders understand the specific bottlenecks that hold teams back, redesign work around AI, and measure whether those changes take hold."
