The hidden challenge in legacy modernisation isn't the legacy code

AI can help organisations understand legacy systems faster, but successful modernisation still depends on understanding the people and processes behind them, says Ben Pirt, Principal Software Engineer at Made Tech, a government transformation specialist with experience modernising complex, large-scale legacy estates.

When organisations think about legacy modernisation, the conversation usually starts with ageing technology. Old code, unsupported platforms and technical debt are all genuine challenges, but they are rarely the biggest obstacle to change.

The harder problem is that legacy systems rarely operate exactly as they were originally designed.

Over years, and often decades, people adapt. They create manual workarounds, introduce new processes and find practical ways to compensate for systems that no longer fully meet their needs. By the time a modernisation programme begins, the technology itself may represent only part of how a service actually works.

That challenge is becoming more widespread. Large volumes of software built in the late 1990s right through to the 2010s are becoming increasingly difficult to support, yet many still underpin critical public services. Replacing them is rarely straightforward. Organisations naturally become reluctant to change technology that keeps day-to-day operations running, particularly where transformation carries the risk of disrupting essential services. The result is often a cycle of delay, with maintenance becoming more expensive while technical debt continues to accumulate.

Much of the public conversation focuses on automation or code generation, but one of AI's greatest strengths is helping teams understand complex systems more quickly. It can analyse applications, extract behaviours from legacy codebases and generate documentation that would previously have taken engineering teams months to produce manually. It can also identify dependencies, support validation and help compare the behaviour of old and new systems during migration.

Taken together, those capabilities have the potential to transform one of the slowest stages of any modernisation programme: discovery.

People remain the missing piece

Legacy systems often sit at the centre of an ecosystem of unofficial processes, manual interventions and operational knowledge that has built up over many years. Teams rely on spreadsheets, informal approval routes and countless local workarounds that exist outside the application itself. In many organisations, the software no longer tells the whole story.

AI can analyse the code, but it cannot explain why someone exports data into a spreadsheet every Friday afternoon before completing a task, or why one team follows a process that nobody documented ten years ago. Those operational realities only emerge by talking to the people who use the service every day.

That distinction matters because successful modernisation is about much more than rebuilding software.

Programmes rarely involve replacing one system with another overnight. More often, old and new platforms must run in parallel while services are migrated gradually, behaviour is validated and risks are carefully managed. That demands engineering judgement, governance and a deep understanding of how services operate in the real world.

AI can accelerate analysis and reduce the effort involved in understanding complex systems, but responsibility for delivering resilient, production-ready services still rests with experienced teams. Rather than replacing engineers, AI allows them to spend less time deciphering legacy technology and more time solving the problems that matter.

There is also a temptation to view AI as a shortcut to transformation. In practice, it is better thought of as another tool in the engineering toolkit. It can make skilled teams more productive and improve the quality of discovery, but it cannot replace the experience needed to modernise critical systems safely or the conversations required to understand how people actually use them.

The future of legacy modernisation won't be defined by AI alone. Organisations that succeed will be those that combine intelligent automation with strong engineering disciplines, operational realism and a genuine understanding of the people and processes that have evolved around their technology over time.

Central Government Programme activities

The techUK Central Government Programme provides a forum for government to engage with tech suppliers. We advocate for the govtech sector, evangelise tech as a solution to public sector challenges, facilitate market engagement, and help make the public sector an easier market to operate in. Visit the programme page here.