The AI adoption gap is no longer access—it is depth of use

ECB survey results show that AI is already present across much of the euro-area business population, but often at low intensity. In the reported distribution, 27% of firms did not use AI, 33% used it very infrequently, 31% used it moderately, and 7% used it significantly.

The gap between awareness and operating depth is now more informative than the gap between users and non-users. Many organisations have crossed the experimentation threshold without redesigning a recurring process around the technology.

Occasional drafting and summarisation can save individuals time, but the gains remain fragmented and difficult to measure. Deeper use emerges when the same method is applied repeatedly, reviewed consistently, and connected to a team output.

This also explains why access alone does not close the adoption gap. Teams need workflow patterns, data boundaries, quality checks, and managers who create permission to change how work is performed.

Segment adoption into four levels: tried, repeated, standardised, and embedded. Then identify one workflow per finance team that can move up one level during the next quarter.

Report adoption through repeated workflow completion and reviewed output—not chatbot messages. Depth is visible when a process becomes faster or more reliable without depending on one enthusiastic individual.