IDC’s latest report published this week, the ‘IDC MaturityScape AI-Fueled Organization 2.0 – 2026 Apr’ and accompanying ‘IDC Agent as Apps-The Rise of Apps – A Vendor Business Model Reset’, captures an important shift in the market. Most organizations are already experimenting with AI. The question is whether they can turn that experimentation into something practical, scalable, and genuinely useful in the day-to-day running of the business.
That challenge is becoming more urgent as the AI landscape gets more crowded. IDC notes that organizations had an average of 40 AI agents in production in 2025, with that expected to rise to 50 in 2026 (source: IDC, IDC’s Future Enterprise Resiliency and Spending Survey, Wave 7, September 2025). That is a clear sign of how quickly things are moving. But it also raises a more important question: what happens when businesses move from a few isolated AI tools into a world of multiple agents, multiple systems, and multiple decisions being made across the operation?
For many businesses, this is where the real challenge begins.
The next phase of AI is not about access to more, but whether AI can actually help the business run better. Can it reduce friction? Can it improve decisions? Can it help teams respond faster, plan better, and execute with more confidence?
That matters especially in manufacturing and distribution.
These are not environments where AI can sit off to the side as a novelty or a disconnected productivity layer. They are environments where everything is connected. Stock affects production. Production affects delivery. Purchasing affects cost and continuity. Service affects customer retention. A delay in one part of the business quickly shows up somewhere else.
That means AI is only valuable if it can work in the reality of the operation.
This is where many organizations are likely to struggle. Not because the technology is not advancing, but because the business itself is often still too fragmented. Critical information sits across ERP, spreadsheets, emails, warehouse systems, shop floor systems, and individual teams. Processes still rely on manual handoffs, workarounds, and people chasing updates between systems that do not fully connect.
AI can sit on top of that. But it cannot fix it on its own.
That is especially relevant in manufacturing and distribution, where a lot of wasted time and avoidable cost still sit between systems, teams, and disconnected steps in the workflow.
IDC’s five-stage maturity model, originally organized around three broad areas: strategy, people, and technology, has been updated this year to elevate governance as a key dimension of AI maturity (source: IDC, IDC MaturityScape: AI-Fueled Organization 2.0, #US54450625, April 2026).
This decision is such an important signal.
As AI moves closer to decision-making and action, governance becomes far more than a compliance issue. It becomes a business requirement. Organizations need to know not just what an AI model or agent can do, but whether it is operating with the right permissions, inside the right controls, and with enough visibility to be trusted.
In manufacturing especially, that matters.
If AI is influencing production schedules, stock decisions, order checks, purchasing activity, customer communication, or anything tied to cost, quality, or compliance, it cannot behave like a black box. Teams need to understand what it is doing, why it is doing it, and whether it can be relied on.
The conversation is also changing in another important way. The real opportunity is in how AI begins to work across the flow of the business.
Some agents will stay narrow and task-based, focused on doing one job well. But the bigger shift comes when AI starts to support a wider workflow or even work across multiple systems and functions. That is where AI starts to move from helping individuals work faster to helping the business run better.
It is also where architecture, context, and governance start to matter most.
That is why one of the most useful recommendations in IDC’s report is its advice to “move agentic experimentation to production by standardizing platforms, integrating enterprise data, enforcing strict guardrails, and scaling proven use cases with clear ownership.” (Source: IDC, IDC MaturityScape: AI-Fueled Organization 2.0, #US54450625, April 2026).
That is ultimately why ERP matters so much in the AI era.
For years, ERP has served as a system of record, capturing transactions, enforcing process, and acting as the operational backbone of the business. That role remains critical. But as AI matures, the role of ERP is beginning to evolve. The systems that create the most value in the next phase will not simply record what happened. They will help organizations understand what is happening, decide what to do next, and act with greater confidence.
That is what the next era of ERP looks like: a system of execution.
And that shift does not happen through AI alone. It happens when AI is grounded in the operational context of the business – jobs, stock, production, purchasing, warehousing, customer demand, supplier dependencies, cost, compliance, and timing – and when it can work inside governed workflows rather than outside them.
This is the future of ERP. Trusted, context-aware execution built on the systems and workflows that already run the business.
That is where the real power of ERP lies in the AI era.
And it is why manufacturers should be thinking not just about where AI can help them move faster, but where it can help them run the business better.
For a broader perspective on how ERP is evolving in the AI era, read Leanne Taylor’s recent article: The AI Execution Imperative.
