Across industries, many organisations are hesitating to act on legacy modernisation, hoping that artificial intelligence (AI) will soon provide a simple, automated fix. It’s an appealing idea, but a dangerous one. Waiting for AI to “solve” problems introduces cost, risk, and uncertainties that grow with every year of delay.
This is particularly prevalent in the banking industry and its continued reliance on legacy systems, such as Gen (formerly CA Gen, COOL:Gen)– which played a key role in modernising the sector in the 1990s but is considered outdated in the age of cloud, data and AI. Gen is still used to build and maintain some of the most critical applications in many of the worlds’ leading banks and insurance companies. It’s also tightly interwoven with other systems and many of the regulatory policies and requirements that must be conformed to.
Banks recognise the need to move away from Gen as costs, risk and uncertainty grow with every year of delay. What’s more, the pool of available Gen skills is diminishing, making the modernisation a business imperative for both continued operation and future innovation.
Go deeper with GlobalData
Access deeper industry intelligence
Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.
The “AI shortcut” is a myth
AI and automation are changing how we approach legacy systems, accelerating discovery, improving documentation, and even supporting elements of code translation. But AI isn’t the transformation silver bullet many are claiming it to be.
The challenges of decades-old Gen applications are structural and strategic – not just technical. These systems are often highly customised and interwoven with other business-critical systems. This means any mistakes whilst transforming Gen-based apps are likely to have a ripple effect on the entire business.
AI can certainly help to modernise Gen, but it is limited. AI can’t make clear, context-driven decisions about architecture, business priorities, or operating models in the same way a professional with decades of experience in Gen can.
The true cost of delay
If banks wait for a future tool to “do it all”, they’ll remain stuck in the past while competitors move ahead. This risk exhibits itself in several ways:
US Tariffs are shifting - will you react or anticipate?
Don’t let policy changes catch you off guard. Stay proactive with real-time data and expert analysis.
By GlobalData1. Rising cost of inaction: Legacy environments are rarely static. Support and licensing costs continue to climb at an alarming rate, and specialist skills become scarcer every year. Even when systems appear stable, the true cost of maintenance, performance tuning, and integration continues to grow.
2. Shrinking talent and knowledge: As experienced Gen specialists retire or move on, their knowledge leaves with them. If banks wait to modernise, capturing and transferring expertise into a modern environment becomes impossible.
3. Operational and resilience risks: Ageing infrastructure is less predictable and more difficult to support. Modern business demands agility, scalability and resilience; qualities that legacy systems were never designed for. When issues arise, recovery times are longer and business impact greater. Some banks are replacing their mainframes entirely, many choosing to migrate these previously unmoveable applications into the cloud.
4. Missed opportunities: Modernisation isn’t only about risk reduction; it’s about enabling growth. Modern platforms open the door to new digital services, improved user experiences, and faster delivery of change. Every year spent waiting is another year of lost opportunity.
Acting now delivers certainty later
It is impossible to approach Gen-based modernisation as a whole. The task of overhauling decades of work is simply too large. But a structured, phased modernisation approach allows banks to build momentum and reduce dependency on legacy technologies while still benefiting from automation and AI where it adds value. There are multiple strategies that banks can use to overhaul Gen-based systems, categorised into five Rs.
Banks could rehost a Gen-based application to another form of infrastructure – physical, virtual or cloud-based – without modifying its code, features or functions. Alternatively, they could refactor the application into maintainable, modern code, meaning applications can be deployed to the cloud, on-premises or hybrid infrastructure of choice. Revising application code is another option, optimising it to remove technical debt left by Gen. As a last resort, banks can rebuild the entire app, redesigning it from scratch and tailoring it to a different environment. The final option is to replace Gen-based apps with alternatives that suit the bank’s current needs.
Each of these strategies have their strengths and weaknesses, and they all reduce cost, dependency, and uncertainty, and position banks to take advantage of future AI capabilities from a position of strength. The right combination depends on individual business objectives, budgets and risk profiles.
Prior to making key decisions, it’s essential that banks conduct a comprehensive but rapid discovery, before seeking to accelerate the migration out of Gen. These crucial first steps enable long-term strategies to be achieved. Getting the application into a maintainable, natural and widely understood coding language can then essentially open the door to various target states. Some banks may even break the application down into functional areas that can be moved to different solutions. But the key is to start on the road to modernisation.
Modernisation can’t come ‘too soon’
Waiting for an undefined technology breakthrough to fix decades of legacy complexity is a costly and uncertain strategy. The banks that act now, understanding their Gen estate, defining business-led priorities, and moving forward in manageable phases will thrive, laying the foundation for innovation and agility.
The real risk isn’t in modernising too soon; it’s in waiting too long.
Chris Eley, Application Modernisation and Transformation Director at TXP
