AI's Role in IT Modernization: Potential and Pitfalls

Rethinking IT Modernization

For many organizations, the journey of IT modernization is a long-standing challenge. Over the years, Chief Information Officers (CIOs) have endeavored to transition applications from older systems, such as mainframes, to more contemporary solutions, including cloud-based environments.

Efforts have included the adoption of microservices, virtualization, and a shift to x86-based servers running Windows and Linux. Despite these advancements, it often seems as though the finish line for modernization remains perpetually out of reach.

Introducing Agentic AI

As organizations reflect on their modernization efforts, new trends are emerging. A recent study by Kyndryl highlights that while many companies invest in modernization, the results have frequently been disappointing. Only about half of the organizations report improved IT operations or significant innovation from their initiatives.

More concerning is that 18% of respondents found little value from their modernization efforts. Furthermore, many organizations are falling behind schedule, facing not only delays but also increased costs.

To turn this trend around, agentic AI technology is stepping onto the scene. Instead of merely replacing outdated systems with newer ones, the emphasis is now on enhancing existing business capabilities.

AI's Impact on IT Ecosystems

Cost savings and escaping legacy systems are no longer the primary drivers for modernization. With the increasing demand for AI across various IT components, business leaders are keen to ensure that their technology landscapes—comprising both old and new elements—function cohesively.

Agentic AI has the potential to intelligently map out the often opaque relationships between legacy and modern systems. Kyndryl’s report noted that fewer than one in ten organizations fully understand these dependencies, leaving a significant technical gap that agentic AI could help address.

Early indications suggest that agentic AI can tackle age-old challenges in modernization processes by streamlining tasks like mapping dependencies, generating code, and creating documentation—tasks that have traditionally required considerable time and effort.

Challenges on the Horizon

Despite the promise of agentic AI, modernization continues to be a complex and ongoing process. As stated by Andy Thurai, founder of The Field CTO and former chief strategist at IBM, it appears that IT modernization is evolving into a chronic endeavor, with many enterprises still relying on mainframes.

Each new technological advancement seems to push the finish line further away. Thurai also notes an addiction to purchasing new software, which complicates the landscape, as legacy systems often remain active rather than being decommissioned.

Furthermore, the introduction of AI can sometimes generate unexpected infrastructure costs, leading to greater expenses without clear benefits. Thurai warns that as vendors increasingly promote AI as a universal solution, leaders may face bewildering costs associated with these new technologies.

The Future of AI in Modernization

Organizations that step beyond superficial AI applications are finding success through the use of agentic frameworks, which help expose hidden technical dependencies. By generating necessary documentation along the way, these organizations are accelerating complex initiatives that previously consumed excessive time and resources.

Ultimately, while agentic AI may appear to be a progressive solution for the challenges of legacy systems, it remains part of an ongoing modernization journey. As Thurai succinctly puts it, today’s AI-driven architectures will likely become the outdated environments demanding future modernization efforts.