The Future of ERP?

I believe we are heading toward a fundamental paradigm shift in enterprise software — and AI is the driver.

Today, ERP vendors are racing to embed AI use cases into their platforms. Copilots, assistants, predictions, automations — all layered on top of the traditional ERP stack. But if you zoom out and look a little further ahead, a more radical question emerges: What if, in the future, there is no ERP system at all? What if there is just AI? More precisely: hundreds, perhaps thousands, of specialized AI agents working together to administer an entire enterprise.

As with every major ERP transformation, it will start with Finance. Imagine a general-purpose AI, supported by dozens of highly specialized agents: one for expenses, one for quotations, one for invoicing, one for incoming payments, one for period close, one for reporting. Together, they handle the full financial lifecycle — continuously, accurately, and without fatigue.

This marks the beginning of the replacement of the human accountant. At first, this will mainly impact very small organizations — companies with fewer than ten employees, sole traders, and freelancers. People who don’t have an affinity with financial administration, or simply cannot afford a full-time accountant. For them, this is not a threat — it’s liberation.

The “AI Accountant” will be delivered as a service, keeping the books for thousands of businesses simultaneously. It will learn at machine speed, improving with every transaction, every exception, every edge case. Over time, it won’t just be good — it will be better than the best human accountant.

And it won’t stop at Finance. More and more agents will be added: Sales, CRM, Procurement, Inventory Management. At some point, calling it an “AI Accountant” no longer makes sense. It has become something else entirely — an AI-driven enterprise operating system.

The AI will learn how different companies price their products and services. It will recognize patterns, successes, failures, margins, and elasticity. And then it will reuse those insights to suggest optimized pricing models for all its clients. It will do the same for taxes.

This is where ethical and legal questions inevitably arise. Is the AI breaching competition law? Is it facilitating tax avoidance — or even tax evasion? For a long time, these questions will remain largely unanswered. Not because they are unimportant, but because policymakers are busy with other priorities — and, frankly, because many of them lack the technical understanding needed to fully grasp what is happening.

As trust grows, the customer base expands. Small businesses become medium-sized. Medium-sized become global. Soon, the first internationally operating companies will adopt these AI-driven systems. The AI will need to understand international trade flows, multiple accounting standards, tax regimes, currencies, and business models — across dozens of countries.

And it will. Because this is exactly the kind of complexity AI is exceptionally good at handling — far better than any human, or any traditional ERP configuration ever could.

Meanwhile, the traditional ERP vendors are not standing still. They are investing heavily in AI, modernizing platforms, and continuing to sell ERP-as-a-Service — understandably seeking returns on the massive investments they made moving their customer base to the cloud.

But the race is on. Is it already too late for them? No one knows. The real question is more uncomfortable — and more urgent: Is their current strategy still the right one?