UK organisations anticipate ramping up their AI investment in 2026, though adoption is still marred by skills gaps, high costs, and shadow AI.
This is according to a new survey by RedHat, which found that UK AI organisations plan to increase AI investment by an average of 32% by 2026.
AI is also the top-ranked priority, along with security, according to 62% of UK orgs, followed by hybrid or multi-cloud (58%), and virtualisation (57%).
However, 89% of organisations surveyed report they are not yet driving customer value from their AI investments.
To overcome these challenges and help turn ambitions to reality, UK organisations are embracing open source across all areas of IT strategy. The survey shows enterprise open source software is considered important for AI strategy by 84% of respondents, for virtualisation by 88%, and hybrid and multi-cloud by 86% and security by 83%.
A Work in Progress
The highest AI priority for respondents (68%) is agentic AI, AI systems that operate with high degrees of autonomy and can execute complex, multi-step tasks with limited human intervention.
After agentic AI, enabling broad employee adoption and operationalising AI are also on the priority to-do list, 68% and 65% agree respectively.
Retaining and developing the right talent remains a challenge, with AI being the highest-ranked urgent skills gap (62% of respondents agree) in the UK for the second year in a row.
Within AI, the talent shortage mirrors the target priorities for UK IT and AI leaders: agentic AI was a top skills gap, cited by 55% of respondents who agree AI has an urgent skills gap. This was followed by the ability to efficiently use AI capabilities (52%) and educating the business overall to use AI (48%).
Almost all (95%) of respondents experience barriers to AI adoption, especially high costs of implementation and maintenance (34%), data privacy and security concerns (30%), and integration challenges with existing systems (28%).
Additionally, 83% of respondents report they are experiencing a “shadow AI” problem – i.e., unauthorised use of AI tools by employees.
Recommended reading
- Data Readiness Severely Lacking for AI Adoption
- Is Data Infrastructure Getting Side Lined in the Push for AI in Health?
- Only 2% of Firms Ready for Agentic AI Risks, New Report Warns
Complex Confidence
Confidence in the UK’s potential on the global AI stage is relatively high, with 83% of respondents agreeing that the UK is a global AI powerhouse or has the potential to become one within three years.
However, this is lower than other nations in Europe, including Spain where the figure is 99%, and Sweden, Germany, and the Netherlands (all 98%).
Respondents cite a lack of talent pipeline (34%), limited public funding (32%), and insufficient private sector engagement (32%) as the main reasons limiting the UK’s rise to AI prominence.
Cloud remains top three on the IT priority list, with AI adding to the complexity as another workload needing to align to evolving cloud strategies.
Barriers to cloud adoption continue: respondents point to internal silos (46% agree), sovereignty concerns (44%), decision pauses on infrastructure investment due to market uncertainty (42%) and unclear ROI (41%).
Drilling down into cloud sovereignty strategy over the next 18 months, UK respondents are prioritising operational control and autonomy (72%), securing the software supply chain (69%) and flexibility and choice of IT suppliers (69%).
“This year’s UK survey results show the gap between ambition and reality. Organisations are investing substantially in AI but currently only a few are delivering customer value.” Joanna Hodgson, country manager, UK, Red Hat, said.
“In the journey from experimentation to sustainable production, enterprise knowledge and integration with enterprise systems must pave the road to achieving value from AI.
“Openness is a force for greater collaboration, sharing best practice and enabling flexibility. As is the case with successful hybrid cloud investments, open source will continue to be the bedrock for making AI more consumable and reusable.”





