Elnion

Analysts at Gartner have identified seven types of system modernization that can be undertaken: Rehost: Lift and shift to a new platform. Replatform: Make minor adjustments to leverage cloud benefits. Refactor: Rearchitect the application. Rebuild: Rewrite the application from scratch. Replace: Swap the application with a new one. Retire: Decommission the application entirely. Retain: Keep the application as-is.

Part 3 of a 5 part series of suplimental commentary blogs following the parent "position paper" recently published titled "Why AI Must Be Measured as Software and Systems not Anthropomorphised Apps" - please read this parent article as a precursor to this 5 part series.. When businesses mistakenly treat AI tools as autonomous staff members or harmless conversational apps, they frequently grant them excessive operational latitude without the requisite IT guardrails.

I've watched organizations wrestle with the question: What should we do about the mainframe? And I often wonder: Why do you need to do anything? There are few topics in enterprise IT that generate as much misinformation as "mainframe modernization." Buzzwords, vendor marketing, and oversimplified narratives have convinced many executives that modernization is synonymous with migration. It isn't. The reality is far more practical and far more valuable.

When organizations discuss reducing IT costs, the conversation often begins with assumptions. Mainframes are expensive. Cloud platforms are cheaper. Legacy systems consume too many resources. Hardware upgrades are the only path to improved performance. In reality, the most effective cost-reduction initiatives rarely begin with infrastructure. They begin with efficiency. For decades, I have worked with organizations running mission-critical applications on IBM Z systems.

When an opportunity to discuss COBOL in a public forum arises, it proves to be an illuminating experience. The 2026 GSE Nordic conference in Helsinki was no exception. Incomparable, or contemptible?

By design, the most important technology in the world is often the least visible. No splashy user interfaces. No viral launch announcements. No breathless headlines about disruption. And yet, every single day you relied on systems running on the IBM Z mainframe. You didn't notice. You weren't supposed to.

Every few years the IT industry rediscovers something that experienced practitioners already know: moving data is difficult. You would think that by now organizations would approach data migration projects with the respect they deserve. After all, the industry has been migrating data since the earliest days of computing… from hierarchical databases to relational systems, from on-premises platforms to distributed environments, and now from legacy infrastructure to the cloud.

Enterprises today are investing heavily in analytics platforms, data lakes, and artificial intelligence initiatives. Cloud scalability, modern tooling, and machine learning frameworks are often portrayed as silver bullets that will finally unlock the value buried in corporate data. Yet despite these investments, many organizations struggle to realize meaningful business outcomes from analytics and AI. Dashboards lag, models underperform, and decision-makers quietly lose trust in the data.

For the global economy, the firewall is the digital equivalent of a castle wall, the fundamental barrier separating order from chaos, legitimate business from sophisticated criminal enterprise. When that wall is breached, the ramifications are immediate and devastating.

For decades now, the “death of the mainframe” has been confidently predicted by vendors, analysts, and industry pundits. Yet, here we are: the mainframe continues to serve as the backbone of mission-critical computing for banks, insurance companies, healthcare providers, retailers, and governments across the globe. It is not only still here—it is thriving. And yet, the drumbeat of “mainframe modernization” never seems to stop.