You are about to spend tens of millions of dollars – possibly hundreds of millions – on an ERP transformation. AI is almost certainly part of the pitch: accelerated implementation, intelligent automation and continuous optimization, among other benefits. The vendors are promising it. Your board is expecting it.
What they are not telling you is that most organizations will fail to capture it. Not because the AI doesn’t work. Because they never properly put it to work.
The Myth That Is Costing You the Transformation
AI arrived in enterprise technology wrapped in a mythology of autonomous intelligence – a system that would figure things out if only given access. That framing has done real damage to ERP programs specifically, because it leads CIOs to treat AI as a capability that gets switched on, rather than a capability that gets built in.
The result is familiar. AI tools are bolted onto the edge of the transformation. The core design decisions – process architecture, data governance, workflow logic, decision rights – get made the old way, and AI is thrown into the periphery. It ends up being used as an advanced search engine, and little else.
That is exactly backwards.
AI performs within the context it is given. Provide it clarity, structure and good data – the hallmarks of a well-designed ERP implementation – and it returns speed, scale and genuine intelligence. Hand it the ambiguity, fragmented master data and undefined workflows that characterize a poorly governed ERP program, and it returns confident errors at volume.
The technology is not the variable. Your implementation model is.
The Onboarding Problem
Think about what happens when a highly capable executive joins your organization. You don’t hand them a laptop on day one and expect results by day five. You invest in context. You share how the business works, what matters, who the key relationships are, what good looks like and where the boundaries of their authority sit. You build in feedback mechanisms. You create space for calibration.
Now consider how most organizations deploy AI in an ERP transformation. No structured context about business priorities or process logic. Vague objectives disconnected from real outcomes. Access to data that is incomplete, inconsistent or ungoverned. No defined boundaries around where AI should act autonomously versus escalate. No feedback loops to sharpen performance over time.
It is the equivalent of hiring your best executive and then refusing to onboard them. You would not accept that outcome with a person. You should not accept it with AI.
The ERP implementations that are generating real returns from AI are doing something deliberate. They are designing AI into the operating model from the start – not adding it on at the end.
What This Means for Your ERP Program Specifically
An ERP transformation gives you something rare, a structural moment to redesign how your organization actually works. The process architecture, the data model, the governance framework – all of it is in motion. That is precisely when the decisions that determine AI’s effectiveness get made.
The organizations pulling ahead are using that moment to make three decisions their competitors are deferring:
First, they are defining where AI acts, where it supports, and where humans decide. Not in general terms. There is specificity by process and transaction type. Which procurement decisions can AI handle autonomously within defined parameters? Which financial exceptions require human judgment? Where does the liability sit, and who owns the override? These are not technology questions. They are governance questions, and they belong in your program design.
Second, they are treating data as the prerequisite, not the parallel workstream. AI in an ERP context is only as good as the master data and its level of granularity, process definitions and business rules it operates within. The organizations that will extract the most value are the ones investing in data governance before they deploy AI capability – not hoping AI will compensate for the data debt they’ve accumulated.
Third, they are building the human side of the human-AI relationship. This means training people not just to use AI tools, but to manage them. Training them to review outputs critically, to give feedback that improves performance over time and to recognize when to override. This is a management discipline. It requires investment in capability, and it requires role design that explicitly accounts for it.
The Force Multiplier You Are About to Buy
AI is one of the most powerful force multipliers available in enterprise operations today. But like any force multiplier, it amplifies what you put in, including your mistakes. A poorly governed ERP program with AI produces poorly governed outcomes faster and at greater scale than a poorly governed ERP program without it.
This is the risk that is underweighted in most transformation business cases. The upside is real. But so is the downside for organizations that treat AI as a feature to be activated rather than a capability to be designed.
The CIOs who will be celebrated for their ERP transformations in three years are not necessarily the ones deploying the most AI. They are the ones who are thinking most carefully right now about how to introduce it, how to integrate it into the workflow and governance architecture they are building. They are thinking how to modify their organization so that humans and AI are each doing what they do best.
The Decision in Front of You
You are making architectural decisions in your ERP transformation today that will determine your AI outcomes for the next decade. Once the process logic is embedded, the data structures are set, and the operating model is locked, retrofitting AI properly becomes exponentially harder.
The question is not whether to include AI in your transformation. That decision has already been made, one way or another. The question is whether you are designing your transformation so that AI can actually perform – with the context, the governance, the data, and the human oversight it needs to deliver on what your board is expecting.
That is an organizational design challenge, not a technology one. And it is yours to own.
Ken Fischer
Ken Fischer is the CEO of Atigro, the proven ERP transformation firm that pairs its modular augmentation capabilities with AI-native frameworks. Atigro’s experience and capabilities generate the rapid development and provisioning of new ERP functionality that meets dynamically changing business processes.
