UK CIOs say AI adoption is outrunning their governance, Logicalis finds

TL;DR:

• 96% of UK CIOs in a new Logicalis survey are worried that proprietary or market-sensitive data could leak through AI vulnerabilities; 95% are worried about losing customer trust through AI errors.
• Only 31% have high confidence in their AI governance frameworks, and just 29% are confident they actively measure the environmental impact of AI initiatives.
• 62% say they are not fully prepared for the EU AI Act and 57% are not ready for the UK’s AI White Paper framework — a number that matters this week, with the EU’s AI Act Omnibus extending HRAIS deadlines into late 2027.

The headline figure — 96% concerned about data leakage — is high enough that any contradiction with adoption pace tells you something about UK enterprise AI maturity. CIOs are simultaneously accelerating deployment and acknowledging they cannot fully see what is deployed: three-quarters say they are only moderately confident they have visibility of the AI tools and services running across their organisations. That is shadow IT, restated for the AI era.

Compliance is the immediate pressure point

Logicalis UKI CEO Neil Eke framed the survey as showing CIOs “under pressure to accelerate AI adoption and deliver value quickly, but governance, visibility and risk controls are still catching up”. The pressure has a specific source: the EU AI Act Omnibus political agreement reached on 7 May extends compliance deadlines for stand-alone high-risk AI systems to 2 December 2027, but the watermarking and transparency obligations still bite from August this year. For the 62% of UK CIOs not fully prepared, the calendar has shrunk faster than the gap has closed.

UK angle: the missing 67%

The deeper finding is structural. Two-thirds of UK CIOs do not have high confidence that their organisations have comprehensive AI risk frameworks and controls — a number that is consistent with the King’s Speech package this week, which framed UK AI regulation around closing the 70%-businesses-using-AI / 7%-using-it-deeply gap. Logicalis’s data suggest that even firms in the “using AI” cohort lack the controls to extend usage safely. That is the actual binding constraint, not access to the tools.

Looking forward

Two expectations follow. First, demand for AI-governance tooling and TPRM frameworks specific to AI deployment will accelerate sharply through Q3 — Logicalis is one of several systems integrators positioning around this. Second, the divergence between the EU’s prescriptive AI Act and the UK’s sectoral-regulator approach is about to become a practical operational issue for UK firms with EU exposure, who are now visibly under-prepared for both regimes.