SAP Customers Want to Know Which Brains Power Their Software
Published on 10/26/2026 at 17:10 | Editorial boerse-global.de
SAP's finance chief has fielded a question that rarely appears on an earnings call: where does the artificial intelligence inside our software actually come from? Dominik Asam told Reuters on Thursday that clients are increasingly probing the provenance of the models embedded in SAP's enterprise products. The disclosure adds a qualitative layer to a cloud story that, on the numbers alone, looks straightforwardly strong.
The answer Asam gave is breadth, not exclusivity. SAP stitches together more than 70 models from a range of vendors, so customer curiosity is about picking within a wide menu rather than reacting to any single supplier. Just as important is what the CFO did not say. No shift in model strategy was announced, no vendor was singled out for removal, and no revenue contribution from the AI features was quantified. The questions point to a demand for information — not yet to proven extra income or a willingness to pay more.
The hard numbers behind the narrative
Those inquiries sit against a set of quarterly figures released on Wednesday. For the third quarter of 2026, SAP reported revenue up 13% to €10.27 billion. Cloud revenue climbed 26% to €6.67 billion, while the current cloud backlog advanced 28%.
The two developments should be kept apart. Overall business momentum is one thing; a dedicated financial proof for the embedded AI models is another, and no causal link between them has been established. Management's guidance reinforces the distinction: for 2026, SAP expects cloud revenue, and combined cloud and software revenue, at the upper end of its existing ranges. That outlook covers those revenue lines specifically — it is not a separate AI forecast.
Should investors sell immediately? Or is it worth buying SAP?
Guardrails before features
Running in parallel is the plumbing that makes agentic AI usable in corporate settings. SAP is folding NVIDIA OpenShell into its Business AI Platform and contributing to the OpenShell codebase, a partnership that goes beyond simply licensing outside technology. The emphasis falls on secure, controllable and auditable execution of AI agents — the conditions under which they operate, rather than a fresh batch of capabilities. No quantified financial contribution attaches to the arrangement; it is a technological step whose worth will be judged by how it converts into offerings customers can actually deploy.
Procurement supplies the concrete use cases. On 20 October, SAP unveiled AI functions for Ariba, Fieldglass and Concur, including the integration of Moody's supplier-risk data into SAP Ariba. A solution built with ORO Labs targets the coordination of procurement workflows. These sit on a different plane from the NVIDIA work: risk data widens the information base, the ORO Labs tie-up smooths processes, and OpenShell governs how agents execute. General availability for the Moody's features is expected in the fourth quarter of 2026 — an expectation, not a completed rollout.
SAP also offers an AI-powered work-intelligence platform, with an agreement in place to expand the technology portfolio; no closing of that acquisition follows from it. The next fixed date on the calendar is SAP TechEd Berlin, including its virtual edition, from 27–29 October 2026.
SAP at a turning point? This analysis reveals what investors need to know now.
For investors, the takeaway is a separation of registers. The cloud business delivers hard growth figures and a guidance frame; Asam's remarks add a customer-side perspective on model provenance; the OpenShell collaboration fleshes out the technical groundwork. Integration, announced availability and economic payoff are three distinct stages — and only the first has clearly begun.
Ad
SAP Stock: New Analysis - 26 October
Fresh SAP information released. What's the impact for investors? Our latest independent report examines recent figures and market trends.
