Outdated legacy systems are continuing to slow the adoption of artificial intelligence and wider digital transformation across the UK public sector, according to new research highlighting the scale of technical debt in government IT estates.
Analysis from Cloudhouse’s State of Technical Debt 2025 report, which surveyed 250 UK IT leaders across government, finance and manufacturing, suggests legacy infrastructure remains a systemic barrier to modernisation across multiple sectors, including government.
The findings indicate that ageing systems are a significant constraint on innovation, with legacy platforms hindering AI adoption for a majority of organisations surveyed. While the study spans several industries, government respondents were singled out for operating within large and complex IT estates that are difficult and costly to modernise.
The report points to widespread technical debt across core operating environments, with 84 percent of government organisations reporting Windows technical debt – slightly lower than manufacturing (92 percent) and finance (89 percent), but still indicative of deep-rooted legacy reliance.
Researchers note that such environments are often maintained to ensure service continuity, but the trade-off is increasing operational risk and reduced agility when integrating newer technologies.
AI ambitions constrained by legacy tech, skills and funding pressures
As government departments accelerate interest in AI-enabled services, integration with legacy systems is emerging as a central obstacle. The report found that three in five organisations say legacy systems are already hindering AI adoption, while a large majority report difficulty integrating AI tools with older infrastructure.
The research also highlights capability gaps as a persistent issue. Most organisations reported lacking the necessary skills to modernise legacy applications, reinforcing concerns long raised in public sector digital strategies about specialist legacy expertise and cloud-native skills shortages.
If you liked this content…
AI in 2026: From proof of concept to public infrastructure
Oracle signs cloud deal with MoD to accelerate legacy modernisation and AI adoption
How to Optimise AI Adoption in Public Sector Organisations
Why SMEs Should Be Central to Government’s AI Ambitions
From Reactive to Agentive: How AI is Redesigning Public Services
At the same time, around 45 percent of respondents said funding is being redirected from innovation towards maintaining legacy systems, slowing cloud transformation and limiting investment in new digital capabilities.
Beyond innovation, the report frames technical debt as an operational and governance issue. Nearly half of organisations surveyed said they had faced compliance challenges in audits due to legacy IT, while outdated systems were also associated with increased downtime and higher cyber risk exposure.
In high-risk sectors including government, where downtime is not an option, this creates additional pressure to modernise without disrupting critical services.
A long-term modernisation challenge
Encouragingly, most organisations surveyed said they plan to modernise infrastructure within the next two years, though only a minority reported that these plans are fully funded.
This gap between intent and investment mirrors broader trends seen across the UK public sector, where transformation programmes must contend with complex estates, procurement constraints and service continuity requirements.
The report concludes that tackling technical debt is not solely a technical exercise but a strategic one, requiring ongoing, incremental modernisation rather than one-off system replacement.
