The UK’s AI race is being slowed by outdated infrastructure, according to new research fromCloudhouse. Its State of Technical Debt 2025 report reveals that nearly 90% of organisations are tied to legacy Windows systems, and 61% say these old platforms are now actively blocking AI adoption.
As the UK heads into 2026 – a year many executives expect to be a critical turning point for AI investment – the report highlights a widening gap between innovation strategy and the operational reality inside IT estates. The survey of 250 UK IT and risk leaders across finance, government and manufacturing shows that while organisations are pushing forward with AI agendas, their foundations may be struggling.
Key findings include:
- 84% say integrating AI into their legacy stack is challenging
- 94% wish more time and budget could be spent on innovation rather than maintenance
- 45% report cloud transformation is being constrained by technical debt
- 48% have already faced compliance challenges during audits due to outdated IT
Looking ahead to 2026, when asked where investment will go next year, AI and machine learning topped the list at 26%, followed by cybersecurity. Yet only 36% say their modernisation plan is fully funded, despite 88% having a two-year roadmap.
This creates a paradox for 2026. Organisations want AI, but the technical debt beneath them is still dominating budget, talent and resources.
More than one in four (28%) finance leaders see AI as a top investment priority - yet finance also diverts more budget (46%) than any other sector into legacy maintenance. On the other hand, manufacturing is the most cloud-constrained, with over half of leaders (53%) feeling the pressure. Government faces the worst skills gap - with 75% lacking the right expertise - and the highest rates of AI projects being blocked altogether, at a staggering 69%.
Mat Clothier, CEO at Cloudhouse, commented: “UK organisations talk publicly about AI ambition, but privately they’re wrestling with systems too old to support it. Technical debt has become the silent hurdle to innovation, draining budgets, exhausting teams and delaying modernisation plans that have already been approved.”
“The UK’s most skilled technologists are spending their days patching, firefighting and sustaining obsolete platforms - rather than enabling the innovation their organisations are prioritising. If the UK wants to lead in AI by 2026, it must first confront the legacy problem beneath the surface,” Clothier added.