Competing in the agentic era from the IBM i platform. The era is not approaching — it is here. This paper is for the technology leader who already knows that, and wants a framework for what to do next.
In early 2026, a major AI lab demonstrated an agentic tool capable of automating dependency mapping, risk identification, and incremental refactoring of legacy COBOL systems. The market reacted as if AI had solved the legacy-code problem outright — that the cost of touching decades-old systems had collapsed, and an era was ending. The conversation that followed was loud, fast, and largely untranslated for the people who actually run these platforms.
That conversation is worth understanding, because buried inside it is a signal that matters directly to anyone running mission-critical infrastructure — or any enterprise codebase with decades of encoded business logic. The press framing was vivid: one widely-read piece called COBOL "the asbestos of programming languages" — once ubiquitous, now dangerously hard to remove. Roughly 95% of U.S. ATM transactions still run on it. Hundreds of billions of lines remain in production. The people who built those systems have largely retired, taking undocumented knowledge with them.
None of those articles were about IBM i. None mentioned RPG. But the dynamics they described — encoded intelligence, aging talent, organizational inertia, and the arrival of AI tooling that can finally reach the logic layer — are not unique to COBOL or to the mainframe. They describe a category of enterprise environment. IBM i is a member of that category.
I have spent decades in that category — across RPG, Db2, DevOps, open source, front-end, and now agentic architectures. So I can say this directly: this is not a future capability. The ability to connect IBM i business logic to an agentic intelligence layer exists today. The tools are here. The architecture is proven. What remains is organizational will and strategic clarity. This paper provides the clarity. The will is yours.
The IBM i platform is the lens for this paper. But the underlying challenge is not unique to it, and leaders who recognize that will take the most value from what follows. A Java shop sitting on fifteen years of custom business logic has the same problem. A .NET shop with a proprietary ERP core has the same problem. A COBOL mainframe environment has the same problem — and now the most press coverage about it.
The encoded intelligence in all of these environments is real, valuable, and largely disconnected from the agentic layer that is rapidly becoming the primary interface for building and operating software. The IBM i case is particularly instructive because the platform pairs exceptional reliability with a purpose-built database in Db2 for i and decades of mission-critical logic running without incident. It is a clear example of the broader pattern: the asset is already built. What's missing is the reachability.
Here is the thesis in one line: the encoded logic of how a business actually runs is the one asset the agentic era cannot manufacture — and the platforms that hold it are positioned, not exposed. Everyone else is trying to teach a model how their business works. The organizations on these platforms already wrote it down, in code, and have been running it in production for decades.
That advantage is real but conditional. It only becomes an advantage if the logic is made reachable — legible to an agent, governed at the boundary, and operable without rewriting the core. The distance between the logic you hold and what an agent can act on is your Knowledge Distance; reachability is how it closes. The platform stays sovereign; the layers above it make it consumable. The work is not migration. It is reachability, with the core protected the entire way.
The path from here is not a single leap. It is a sequence, built in order, on a foundation that can govern it. Governance, security, and auditability first — the controls that make opening the core to agents safe. Then the core made reachable through APIs without changing it. Then the layer that makes it agent-legible. Then agents standing up thoughtfully, in tandem with the organization's own readiness. You prove it on one slice, then scale, then run fully — always on the same foundation, always with the core protected.
What makes this navigable rather than overwhelming is that you do not have to do all of it at once, and you do not have to bet the business to begin. The discipline is sequence: build each layer before the one that depends on it, so the governance layer never falls behind the execution layer. That ordering is the difference between agentic capability that holds and capability that breaks the first time it meets real traffic.
The posture is not urgency for its own sake, and it is not waiting for the dust to settle. It is the posture of an operator who understands that the advantage is real, the window is open now, and the work is buildable — and who chooses to build it deliberately rather than react to it late. The platforms that hold the world's encoded business logic are not the ones at risk in the agentic era. They are the ones with the most to gain, if they move with clarity.
That clarity is what this paper exists to provide. The rest is execution — and execution is where advantage is actually kept.