Welcome to this month's ERP news roundup. We track what is actually changing for ERP buyers - not every vendor press release, but the shifts that affect how real businesses choose and run their software. The headline for the past month is impossible to miss: the three biggest names in enterprise software all put the same phrase at the center of their messaging. SAP talks about the "Autonomous Enterprise," Oracle shipped "NetSuite Next," and Microsoft rolled out another expansion of its Dynamics 365 agents. Strip away the branding and they are all describing the same thing - ERP that embeds AI agents directly inside business processes, rather than treating AI as a bolt-on assistant.

This matters because when SAP, Oracle, and Microsoft converge on the same strategy in the same year, it stops being a bet and starts being the direction of the whole category. For SMEs, the practical questions are not "is agentic ERP real?" - it clearly is - but "what is actually shipping today versus what is still a demo," "what will it cost," and "where should a smaller company start." Here is what landed, and what it means.

Microsoft Expands the Dynamics 365 Finance Agent

Microsoft's most recent Dynamics 365 update, published in late September 2026, widened the scope of its Finance Agent considerably. Where earlier versions focused on a narrow set of tasks, the agent now reaches across reconciliation, data preparation, variance analysis, collections, and business intelligence - and, importantly, it is accessible from where finance teams already work: Copilot, Excel, Outlook, and Teams, rather than only inside the ERP's own screens.

That last point is the quietly significant one. A reconciliation agent that lives inside Excel, where the accountant is already building the workbook, removes a context switch that kills adoption. The lesson from a decade of ERP automation is that tools people have to go out of their way to open tend to go unused. Meeting finance teams in their existing tools is a meaningful shift in how this software gets adopted, independent of how clever the underlying model is.

SAP's "Autonomous Enterprise" and Oracle's NetSuite Next

SAP has framed its 2026 strategy around the "Autonomous Enterprise," with its Joule assistant and a large and growing catalog of agents spanning S/4HANA, SuccessFactors, and its Business Technology Platform, plus a generative-AI layer customers can configure with their own data. The direction is clear: dozens of assistants and hundreds of agents aimed at taking routine work off people's desks across finance, HR, and operations.

Oracle answered with NetSuite Next, bringing conversational, agentic workflows into its mid-market ERP and bundling a large set of AI enhancements into the product. A notable difference has emerged in how the vendors price this. Oracle has leaned toward treating many of its AI enhancements as standard, included capabilities, while SAP has positioned at least some of its advanced AI as a premium tier that carries additional cost. We would caution against reading too much into any single pricing comparison - packaging, editions, and discounts vary enormously by contract - but the strategic contrast is worth watching: "AI included" versus "AI as a premium add-on" is becoming a real axis of competition, and it directly affects the total cost a buyer should budget for.

This month in one sentence

Agentic ERP stopped being a keynote demo and became a shipping product line at all three major vendors - which means the real questions for buyers are now about scope, governance, and price, not whether the technology exists.

How Big Is This, Really? The Numbers Behind the Hype

It is easy to get swept up in launch events, so it helps to anchor on the analyst view. Gartner expects roughly 40% of enterprise applications to include task-specific AI agents by the end of 2026, up from under 5% in 2025. That is a genuinely steep adoption curve - but notice the phrase "task-specific." The near-term reality is a growing library of agents that each handle a bounded job (matching an invoice, chasing an overdue payment, flagging an anomaly in the ledger), not a single system that runs the finance function unattended.

That distinction is the whole game for a smaller company. "Task-specific agent with a human approving the result" is available, useful, and governable today. "Fully autonomous ERP" is still mostly a roadmap. Vendors that are honest about which is which will save their customers a lot of disappointment. Organizations that have integrated AI into their ERP workflows have reported real operational gains - shorter cycle times and lower processing costs on the tasks they have automated - but those results come from disciplined, narrow deployments, not from flipping an "autonomous" switch.

What This Means for SMEs

If you run or advise a growing business, the arrival of agentic ERP at the big three does not mean you need to rip anything out. It does mean a few things are worth doing this quarter:

  • Separate shipping features from roadmap promises. When a vendor demos an agent, ask directly: "Is this generally available today, and can I talk to a customer running it in production?" The gap between announced and in-production use cases is still wide.
  • Budget for the AI pricing model, not just the licence. Ask whether AI capabilities are included in your edition or priced as a premium add-on, and get the all-in number before you compare vendors.
  • Start narrow and keep a human in the loop. Pick one high-volume, low-ambiguity task - bank reconciliation or invoice matching are the usual starting points - and keep a person approving the agent's output until it has earned trust on your real data.
  • Ask where the agent runs and what it can touch. Agentic features raise new governance questions - which we cover in a companion piece this month. Make sure you understand what data an agent can access and act on before you switch it on.

Where Inovexa fits into this trend

We build Inovexa for SMEs, so we watch the autonomous-enterprise race with a specific question in mind: which of these capabilities actually helps a 20-to-200-person company, and which are built for enterprises with a dedicated automation team? Our approach is to embed AI where it removes genuine drudgery - reconciliation, matching, first-draft reporting - with the human firmly in control of anything that touches the books.

We are not going to tell you AI runs your whole business - nobody's does yet, honestly. What we can do is walk through which agentic features are mature enough to rely on today and which are worth waiting on, for a company your size.

The Bottom Line

The past month confirmed that agentic ERP is now the shared strategy of SAP, Oracle, and Microsoft, not a side experiment. For the enterprise, that means an arms race of assistants and agents. For SMEs, the smart move is less dramatic but more valuable: understand what is genuinely shipping, insist on clear pricing, and adopt one well-governed agent at a time. The companies that win with this technology will not be the ones that switched everything on first - they will be the ones that chose the right few tasks, kept their controls tight, and expanded from a position of trust.

Talk to our team if you want a straight answer on which agentic ERP features are ready for a business your size.

More ERP news this month: AI Governance Becomes an ERP Buying Criterion · Agentic AI Reality Check: Where SMEs Should Start

Further reading & sources: Microsoft - Build the future of agentic ERP with new Dynamics 365 capabilities · Gartner - ERP Insights · Forrester - Predictions 2026: AI Agents & Enterprise Software.