AI in IBM i Modernization: A Strategic Assist, Not a Shortcut

If you’ve attended a tech leadership meeting in the past year or so, you’ve probably heard this:

“AI is going to change everything, from Development to Modernization, even how we approach long-standing systems like IBM i.”

That’s encouraging. But it also leads to a simple question: What exactly is AI going to change?

For many IBM i teams, the pressure is real: RPG expertise is retiring, integration demands are rising, and modernization can’t afford business disruption. AI is becoming valuable not because it replaces engineers, but because it reduces uncertainty early.

When it comes to IBM i modernization, the answer is more down-to-earth than the buzz. AI doesn’t get to choose your architecture. It doesn’t substitute for good engineering judgments. And it certainly doesn’t rewrite decades of business logic overnight.

What AI does is allow people to understand their systems better. It reveals underlying logic, dependencies, and structure much faster than traditional discovery processes. In this way, AI is a helper, not a replacement.

That’s an important point. Modernization succeeds or fails based on understanding, first. The rest is up to humans.

How AI Helps in IBM i Modernization

The greatest value of AI isn’t what it replaces. It’s in what it helps you understand faster.

AI doesn’t eliminate the need for skilled teams, but it can reduce the manual effort required for discovery, documentation, and testing.

1. Comprehending Decades-Old Logic

If you’ve ever worked on an outdated IBM i platform, the hardest part is figuring out what the platform does. Not a single guide is available. Old RPG or COBOL code, some of which hasn’t been updated in years, contains the business logic. In some cases, the individuals who were aware of it have passed away.

For example, an order-entry RPG program may contain pricing, tax, inventory, and invoicing logic all intertwined, making modernization risky without clarity.

In this regard, AI can serve as a backstage helper.

AI-assisted discovery tools can summarize legacy business rules, surface hidden dependencies, and generate usable documentation in weeks instead of months. Some of them even provide a visual representation of the logic in the form of a diagram or a summary that doesn’t require line-by-line analysis.

2. Mapping Dependencies

IBM i environments often have configurations that were not designed for modularity. Over the years, they became interwoven. Change one file, and another might fail.

AI-powered mapping solutions can help make sense of this.

By analyzing code and data flow, they reveal how screens, jobs, files, and programs are related, even if it’s not immediately clear. This includes green-screen workflows, CL batch jobs, nightly billing runs, and tightly coupled DB2 file structures. You get a clear picture of what’s active, what’s idle, and what relies on something else. This makes impact analysis easier and less surprising during modernization.

3. Testing Scenarios Before You Commit

Now that you have a good understanding of the existing system, it’s time to make changes. But changes are uncertain.

What if we break the monolith into services? What if we refactor the UI? Will the downstream logic still work? Will the integrations remain intact?

AI can support impact analysis and regression test generation, helping teams identify where changes could break downstream workflows before deployment.

They won’t provide a definitive answer (that’s up to design), but they let you test your hypothesis. If a change is going to slow things down or mess up the logic, the solution points this out.

You get a sneak peek at what could go wrong before it actually does. This is very useful, especially when trying to meet modernization goals while keeping business continuity.

Where AI Still Falls Short

AI is fast and powerful, but it has its limitations, especially when it comes to big, strategic decisions. It can assist you in understanding systems, but it cannot make decisions, change team culture, or assume risk. That’s still on us.

This is where humans are still doing the heavy lifting:

I. The Important Choices Regarding Your Architecture

Rebuild or replatform? Do you want to start over or keep what you have? These are business decisions as well as technical ones. Budget, talent, time, and organizational adaptability are all important factors. There isn’t an AI solution that can read the room and balance these factors.

The CIO, the architects, and the business executives who know what the company is prepared for make this decision. Code is only one aspect of modernization. Risk, timing, and what the company needs next are all important considerations.

II. Aligning Individuals

No amount of AI genius can solve problems with communication between departments or between product and infrastructure teams with disparate motivations. Nothing will save the project if teams are not working together.

Research consistently shows most transformations fall short due to organizational alignment, not tool readiness. When people aren’t aligned, even the best tools struggle to make an impact. At its core, successful change is less about systems and more about shared ownership.

III. Owning the Risk

You have to decide whether to go live or stick with this plan at some point during modernization. AI will not be in the room for this either. It doesn’t address outages or provide an explanation for why SLAs weren’t fulfilled. The project managers are still in charge of that.

Therefore, while AI can be used for speed and insight, it should not be confused with ownership. It is necessary for someone to verify what has been constructed, test potential problems, and prepare for potential setbacks. Only people can be held responsible when the stakes are high.

IV. Governance, compliance, and explainability

AI cannot replace audit controls or validation in regulated environments.

The Mindset Shift: From Automation to Augmentation

For a long time, people have been talking about AI as automation in modernization: fewer people, faster decisions, lower-cost outcomes.

This sounds good, but it can be very frustrating.

The truth about AI is augmentation: it assists teams in working smarter, not harder. The key to success is to think of AI as a teammate, not a replacement for human intelligence.

The most successful modernization projects are those that simply integrate AI into discovery and documentation, providing input while retaining human control.

How Damco Applies AI in IBM i Modernization

At Damco, we use AI to accelerate IBM i discovery, dependency mapping, and refactoring support, while modernization remains phased, governed, and business-led. Our focus is modernization without disruption: improving agility, enabling integration, and reducing long-term risk without rewriting what already works.

A Smarter Way to Start

If you are considering modernization with AI, start in a smarter way.

1. Don’t Chase AI. Chase Friction

Identify areas within your team that are causing friction today. What is taking too long? What is being reworked? What documents are missing?

This is your AI use case.

2. Start Small, But with Impact

Start with one module, not the entire system. Use AI to analyze logic, visualize dependencies, or simulate changes and compare outcomes before and after.

If it saves just a few hours a week, that’s a step in the right direction. Start there.

3. Keep Humans in the Loop

Don’t try to automate indiscriminately. Use AI where it accelerates understanding but keep humans accountable for decisions.

Think of AI as your analyst, not your architect.

Final Thought: AI Doesn’t Modernize Your Systems. People Do.

There’s no question that AI is making IBM i modernization faster and smarter. It shortens the path to discovery. It reveals logic that used to be buried in layers of legacy code. It helps you spot risks before they become real problems.

But AI doesn’t make decisions.

It won’t map your modernization journey. It won’t reorganize your teams. It won’t explain to management why something succeeded or failed.

That’s left to people.

Business decisions have technical implications, not the other way around. The most effective companies aren’t racing to adopt AI for its own benefit. They leverage AI to build insight, then make thoughtful, human decisions based on what matters to their business.

So AI can help you go faster. But sound thinking precedes control.

In a large-scale modernization project, sound thinking is the only benefit that will continue to compound.

If you’re trying to determine the role of AI in your IBM i strategy, begin with this question: What do we need to understand better before we make any changes?

That’s where true progress starts.