AI is changing the way businesses think about their technology strategies. Generative AI, intelligent automation, predictive analytics, and AI agents are creating new opportunities to improve efficiency, but their impact goes deeper than that. At their core, these technologies are changing how organizations access, understand, and act on information.
As interest in AI grows, deciding where to start can be challenging. While adding an AI tool to an existing application or adopting a new model may seem like a logical first step, that only addresses part of the equation. For AI to deliver meaningful results, it needs reliable access to an incredibly valuable asset: data.
For organizations running critical workloads on IBM i (the modern evolution of AS400), this brings the AI conversation back to a more practical starting point. The question is not whether IBM i is modern enough for AI and intelligent workflows, but how organizations can bring these technologies into an established environment without disrupting critical operations—a widely-recognized issue in the IBM i space.
Modernization helps bridge that gap. By making IBM i data more accessible to modern applications, analytics platforms, cloud services, and AI technologies, organizations can create a foundation that extends well beyond a single AI initiative.
Viewed this way, your AI strategy is really a data strategy. Let’s take a closer look at how IBM i organizations can build that foundation, including important considerations like data accessibility, business context, and governance.
Does IBM i Have the Data I Need for AI?
AI depends on data, but simply having large amounts of information is not enough. To make your AI useful, any new tools or systems require relevant business data, as well as enough context to understand what that data means and how it relates to the business.
IBM i environments can be rich sources of both. As a core platform of many high-volume industries such as banking, insurance, manufacturing, retail, logistics, and others, IBM i can hold tons of valuable information gathered from disparate parts of the business. Put short, these systems may support processes involving customers, transactions, inventory, orders, financial information, production, and other essential areas of the business. Depending on the processes you are trying to improve with AI, this context is incredibly important.
This creates an opportunity to think about IBM i differently within an AI strategy. Rather than assuming information needs to be moved away from IBM i before it can become useful, organizations can focus on making the data and business capabilities they already trust more accessible to the rest of the enterprise.
With this in mind, it’s important to note that valuable data is not automatically usable data. Making that data usable is where the modernization challenge begins.
AI & Data Accessibility: The IBM i Integration Challenge
Consider an organization introducing an intelligent customer service application. The application may need to work with customer account information, recent orders, inventory availability, pricing, or transaction history. Some of that information could reside in a CRM or cloud platform, while other pieces may be stored on IBM i.
The organization already has the information it needs, but accessing it in a secure, consistent, and useful way may be more complicated. A few common challenges include data scattered across systems, applications tied closely to long-standing database structures, and integrations built for one process rather than broader reuse. Batch-oriented exchanges can also fall short when newer use cases require faster access to information.
While adding AI does not necessarily create these architectural challenges, it can make them much more visible. Traditional applications are often designed around well-defined inputs, outputs, and workflows, but AI, analytics, and intelligent automation can introduce new expectations for how information moves between systems. This can quickly highlight just how easily (or not) you can access data across different applications.
Security and governance add another dimension. Connecting an AI application to enterprise data should not mean providing unrestricted access to that information. Organizations still need appropriate controls over what data can be accessed, which applications or users can access it, and how that information is used.
As AI becomes more closely connected with operational processes, the architecture supporting the data becomes increasingly important. That is where AI readiness and IBM i modernization begin to overlap.
What Does Modernizing for AI Look Like?
Modernization can take many forms in an IBM i environment, from updating RPG applications and database structures to adopting modern SQL, APIs, services, and new development practices. Even for AI initiatives, the focus does not need to be on replacing IBM i. It is often about targeting specific business capabilities and how they connect with the rest of the enterprise.
Instead of building a custom extract or point-to-point connection every time an AI application needs IBM i data, organizations should establish reusable, governed methods of access. Those connections can serve AI while also supporting analytics, automation, cloud applications, and other enterprise workflows.
For example, an organization may have reliable IBM i applications built around database structures and access patterns established years ago. Those applications can continue serving the business while the architecture around them evolves. This can include:
- Modern SQL and SQL views to provide cleaner access to Db2 for i data
- REST APIs to expose IBM i data and business functions to modern applications
- Off-the-shelf or custom middleware to connect IBM i with cloud and enterprise platforms
- Data pipelines to move and prepare IBM i data for analytics and AI use cases
- Event-driven integration to support workflows that need more timely data exchange
- Targeted application refactoring to make specific applications or business functions easier to integrate
The exact approach will depend on your existing architecture, security requirements, data needs, and use case. The goal should be to make IBM i easier to connect with whatever you’re building next. When modernization creates that flexibility, it creates a foundation that you can build on for years to come.
Contextualizing Your Data for AI
Now onto the next step in your strategy. Modernizing IBM i to make data easier to access is an important step, but accessibility alone does not make it ready for AI. The information also needs to be accurate, consistent, and give AI enough context to use it correctly.
This matters because enterprise data rarely exists in isolation. A customer number, product code, transaction status, or inventory value may make perfect sense within the IBM i application that created it, but an AI system needs enough context to interpret that information correctly when it is combined with other sources.
Businesses can address this by preparing the data around the needs of each AI use case. This may include (but is not limited to):
- Defining important data fields and business terms so that applications understand what values, codes, statuses, and abbreviations mean.
- Mapping relationships between systems so that data can be correctly connected with records from CRM, cloud, analytics, or other platforms.
- Documenting relevant business rules when the meaning of data depends on application logic or other established processes.
- Standardizing and validating data to address duplicate records, inconsistent formats, missing values, or conflicting definitions.
Long-running IBM i environments can make this work especially valuable. Database structures, field names, application logic, and business rules may have evolved alongside the business for years.
The challenge is that this knowledge is not always well documented. Important relationships, dependencies, and business rules can be embedded in RPG or CL source code, while the people who originally built or maintained an application may no longer be available to explain it. Tools such as Programmers.io’s Impact Analysis can help fill these knowledge gaps by analyzing IBM i applications to uncover dependencies, extract business rules, and generate technical and functional documentation.
Modernization also provides an opportunity to make some of that knowledge more explicit and usable outside of the applications it originated from. AI can deliver greater business value when it can access the right information, understand how it connects, and apply it in meaningful ways.
Beyond Implementation: Approaches for Governing AI Responsibly
Making IBM i data accessible and understandable are important parts of an AI data strategy, but access also introduces risk. As data becomes available to AI systems and automated workflows, organizations need to consider what AI should be allowed to do with that information. Without clear safeguards, organizations risk exposing unnecessary information, using data outside its intended purpose, or losing visibility into how AI-supported decisions are made.
A practical place to start is with a small-scale use case. Companies should identify the data the AI needs to function appropriately, determine which systems and users should have access to it, and apply robust security, privacy, auditing, data-quality, and human-oversight requirements before expanding access.
This type of governance is becoming increasingly formalized as enterprise AI adoption grows. ISO/IEC 42001:2023, for example, establishes an Artificial Intelligence Management System focused on how organizations govern AI and manage related risks such as transparency, ethical considerations, security, continuous adaptation, and others.
As part of your modernization roadmap, investing in security measures aligned with your specific use case and regulatory landscape not only helps mitigate security and compliance risks, but creates a stable foundation for expanding AI adoption in the future.
Getting Started
A successful AI strategy starts long before choosing a model or launching a new tool. It starts with the data those technologies depend on and whether that data is accessible, understandable, and appropriately governed.
For IBM i organizations, modernization provides a practical path forward without requiring a risky migration of your critical operations. By improving how IBM i data is accessed, integrated, contextualized, and protected, businesses can create a stronger foundation for AI while supporting future initiatives.
Your AI strategy may start with a specific business need, but ultimately, the strength of that strategy will depend on the data foundation underneath it.
If you’re exploring how AI can fit into your IBM i environment, Programmers.io can help you identify the right data, integration, and modernization priorities to move your strategy forward. Our IBM i experts are ISO42001-certified and have the experience to help you navigate enterprise transformation safely, securely, and without disruption. Connect with an IBM i expert today to get started.
