A benchmarking report from the Business Transformation Europe Summit has found that while 70% of organisations describe AI as critical to their strategic goals, only 11% are systematically using it to deliver transformational outcomes. The findings, drawn from a survey of more than 200 transformation leaders globally, suggest the central challenge for enterprise AI has shifted from the question of whether to invest to the harder problem of how to scale.
The report, which carries contributions from practitioners at Google and the University of Pennsylvania alongside data from the 2025/26 PEX Report on the Global State of Business Transformation, is positioned by its publisher as a benchmarking tool for executives trying to measure their organisation’s progression from isolated pilots to enterprise-wide deployment.
What the gap signals
The 59-percentage-point divergence between stated strategic intent and measurable delivery is not a new phenomenon, but the scale of it underscores a persistent structural problem. Organisations frequently invest in AI proofs of concept that demonstrate value in narrow, controlled conditions and then stall at the point of scaling. Contributing factors typically include fragmented data infrastructure, unclear ownership of AI governance, and the absence of operating models designed to absorb AI outputs into existing workflows.
The report frames this as moving from “random acts of innovation” to what it describes as “cultivated bouquets,” a metaphor for deliberate, coordinated AI deployment across business functions rather than ad hoc experimentation.
Governance and operating model as the real bottleneck
For fintech organisations specifically, the governance question carries additional weight. Firms operating under FCA oversight, or subject to DORA requirements in the EU, face a regulatory expectation that AI systems used in consequential decisions, including credit assessments, fraud detection, customer communications and compliance monitoring, are explainable, auditable and subject to meaningful human oversight. Scaling AI without a governance framework that satisfies those expectations is not simply an efficiency risk; it is a regulatory one.
The broader European market is working through this tension.