Mainframe Skills Aren’t Disappearing; They’re Becoming More Valuable

Craig Mullins explains how mainframe skills are indispensable in an AI-driven world as mainframers assume a more strategic role

Craig Mullins February 19, 2026

Every few years, a familiar prediction resurfaces: This will be the year the mainframe finally fades away. Along with it we usually see a companion claim that mainframe skills are obsolete, difficult to staff and destined to vanish as workloads move to the cloud.

And yet, the IBM Z platform continues to run the world’s most critical systems. Banks clear transactions in milliseconds. Airlines manage global reservations. Governments process taxes, benefits and national records. These workloads are not shrinking. They are growing in scale, complexity and importance.

Indeed, this is backed up by IBM’s most recent quarterly financial results that showed a 67 percent year-over-year gain in IBM Z revenue. IBM attributes this growth to heightened demand for its latest z17 mainframe.

What is disappearing is not the mainframe, nor the need for skilled professionals, but rather an outdated view of what mainframe expertise looks like. Far from losing relevance, mainframe skills (especially in areas like Db2 for z/OS) are becoming more strategic and more valuable than ever.

The Myth of ‘Legacy’ Skills

The term “legacy” is often used carelessly. In many organizations, it has become shorthand for “old” or “replaceable.” But in technology, longevity is often a sign of success, not failure. Many IBM Z practitioners have transitioned from the term “legacy,” replacing it with “legendary,” feeling it more aptly describes the platform. And I agree!

For example, Db2 for z/OS has endured because it consistently delivers what enterprises need most: availability, integrity, scalability and security. These qualities are not less important in a cloud- and AI-driven world; they are foundational.

What has changed is the context in which these systems operate. Mainframe data is no longer isolated. It feeds mobile apps, analytics platforms, AI models and distributed services. This interconnected reality requires more skill, not less, from the professionals who manage it. This includes traditional skills such as ensuring the performance and availability of Db2 databases and applications, but also newer skills such as integrating mainframe data and platforms with cloud and distributed environments.

From Operators to Strategists

However, the traditional image of the mainframe professional—someone who “keeps the lights on” by managing batch jobs and monitoring system health—no longer reflects reality. Today’s mainframe specialists are increasingly:

Data stewards, ensuring trusted data is shared safely across hybrid architectures
Performance engineers, balancing mixed transactional and analytical workloads
Automation designers, defining guardrails for AI-driven and policy-based tooling
Risk managers, protecting mission-critical data from security, compliance and availability threats

In other words, working on mainframe systems has shifted from being operationally focused to more strategic.

This is particularly evident in the evolving responsibilities of the Db2 DBA.

The Modern Db2 DBA Skill Set

The idea that a DBMS such as Db2 for z/OS can “run itself” is appealing but dangerous. Automation, and increasingly AI, can streamline routine tasks, but it does not replace judgment, experience or architectural understanding.

Modern Db2 DBAs must understand:

Dynamic SQL behavior and access path stability
Workload isolation and performance predictability
Data sharing and continuous availability architectures
Integration with distributed analytics and AI platforms
The operational impact of autonomous features and AI-assisted tuning

As automation increases, the consequences of mistakes grow. A poorly governed AI recommendation or an unchecked automated change can ripple across thousands of transactions per second. Skilled DBAs are the ones who understand when to trust automation and when to override it.

AI Increases the Value of Trusted Data

Instead of eliminating the need for mainframe skills, the rise of AI has amplified the importance of mainframe expertise.

AI systems are only as good as the data they consume. Enterprises are quickly discovering that their most accurate, complete and trusted data often resides on IBM Z. That makes the mainframe a critical participant in AI initiatives, not an obstacle to them.

But exposing mainframe data to AI workloads introduces new challenges, including:

Data consistency and timeliness
Governance and access controls
Performance isolation between operational and analytical usage
Protection against unintended data misuse

These are not problems that can be solved by tools alone. They require professionals who understand the data, the platform and the business implications of every design decision.

Skills Shortage or Skills Transition?

Much has been made of the “mainframe skills shortage.” In reality, the challenge is not a lack of capable professionals. Instead, it is a skills transition problem.

Organizations that struggle are often those that:

Treat mainframe teams as isolated from modernization efforts
Fail to invest in training beyond basic operational tasks
Do not promote collaboration and skills development at industry events
Assume automation eliminates the need for deep expertise

Conversely, organizations that thrive are integrating mainframe professionals into broader architecture, data and AI conversations. They are training experienced staff in modern tooling while teaching newer professionals the fundamentals of reliability, performance and scale that IBM Z exemplifies.

Younger technologists are not opposed to the mainframe. They are opposed to being siloed. When mainframe roles are positioned as strategic, modern and impactful, interest follows.

Why These Skills Command a Premium

As enterprises move toward hybrid and AI-enabled architectures, professionals who understand both traditional reliability engineering and modern data integration are rare but also increasingly valuable.

Mainframe expertise sits at the intersection of:

High-volume transactional processing
Enterprise-grade security and compliance
Data governance and integrity
Hybrid cloud integration
AI and analytics enablement

That combination is difficult to replace and expensive to relearn the hard way. As a result, organizations are discovering that retaining and developing mainframe talent is far more cost-effective than attempting wholesale replacement.

The Future Is Not Less Mainframe Skill; It’s Better Mainframe Skill

The mainframe is not standing still, and neither are the professionals who support it. Tools will continue to evolve. Automation will continue to improve. AI will increasingly assist with tuning, monitoring and analysis. But the need for expertise … that is, real expertise … will not diminish.

If anything, the stakes are higher. When platforms underpin global commerce and feed enterprise AI systems, mistakes are not merely inconvenient; they are existential.

Mainframe skills are not disappearing. They are being reframed, elevated and integrated into the heart of modern IT strategy. The organizations that recognize this will not only preserve their critical systems; they will turn them into competitive advantages.

And the professionals who master this new reality will find that their skills are not only relevant, but indispensable.

Tips and Resources: Where to Acquire Modern Mainframe Skills

If you believe that mainframe skills are going to continue to be important, and better mainframe skills will be valued, then it stands to reason you will want to acquire new mainframe skills and improve your existing ones. But how can that be done?

The perception that mainframe skills are hard to acquire is outdated. In reality, there are more learning paths available today than ever before. And importantly, many of them are designed to blend traditional IBM Z fundamentals with modern development, data and automation practices.

IBM Training and Digital Learning

IBM offers a broad catalog of self-paced and instructor-led courses covering IBM Z, z/OS, Db2 for z/OS, security and performance topics. These courses increasingly emphasize automation, hybrid integration and real-world scenarios rather than purely operational tasks.

IBM Z Xplore

Instead of just static coursework, IBM Z Xplore emphasizes experiential learning by offering hands-on challenges, guided learning paths and real-world scenarios built around modern IBM Z technologies. You can gain practical experience with tools, programming and system concepts that go well beyond theory.

IBM Z Academic Initiative

Universities and colleges worldwide participate in the IBM Z Academic Initiative, giving students hands-on access to real mainframe environments. This program has become a primary on-ramp for early-career professionals entering the IBM Z ecosystem with practical experience.

Vendor and Tool-Based Education

Independent software vendors in the IBM Z space often provide specialized training focused on performance management, recovery, security and automation. These offerings are particularly valuable for experienced professionals looking to deepen expertise in specific operational or architectural areas.

Community and User Groups

Organizations such as SHARE, GSE, IDUG and regional mainframe user groups remain invaluable sources of education. Conferences, webinars and technical sessions provide insights that go beyond documentation—often delivered by practitioners solving real production problems.

On-the-Job Mentoring and Apprenticeship

Despite the availability of formal training, nothing replaces guided experience. Organizations that deliberately pair newer staff with seasoned professionals accelerate skills transfer and reduce operational risk. Structured mentoring remains one of the most effective ways to build deep mainframe expertise.

Modern Skills Matter, Too

Today’s mainframe professionals benefit from knowledge beyond traditional boundaries. Familiarity with SQL optimization, APIs, DevOps concepts, automation frameworks and data governance enhances effectiveness and relevance in hybrid environments.

The most successful learning paths combine formal education, hands-on practice and real-world problem solving. Mainframe skills are not acquired overnight. But for those willing to invest, they offer long-term relevance and significant career value.