Why SAP Put Its CEO in Charge of AI—and What Must Happen Next

SAP’s AI push is no longer simply a product expansion. It is a test of whether the company can convert its position at the center of enterprise finance, supply chain, procurement and HR into an enduring advantage in agentic AI.

That explains the urgency behind the July 2026 leadership reshuffle. SAP has a credible architectural proposition—the Business AI Platform, Joule and the Autonomous Suite—but market pressure and questions about the pace of delivery mean it needs visible customer results quickly. 12

Why the supervisory board wants a breakthrough

Bloomberg reported that members of SAP’s supervisory board believe the company needs an AI breakthrough in the coming months to avoid falling badly behind competitors; more optimistic directors put the window at roughly two years. 2

The concern is not that SAP lacks a large installed base or cloud business. It is that investor expectations have shifted toward proof that enterprise software providers can deliver practical AI automation.

SAP’s January 2026 results illustrate the pressure. The company reported full-year 2025 cloud-revenue growth of 26% at constant currencies and total cloud-backlog growth of 30%. But its 2026 cloud outlook fell short of market expectations, and shares dropped 15% on January 29—the company’s steepest one-day decline since October 2020. Current cloud backlog grew 16% as reported in the fourth quarter, or 25% at constant currencies, to €21.1 billion.

Those results did not erase SAP’s underlying cloud progress. They did, however, make growth durability and AI execution much more important to investors. In August, UBS downgraded SAP to Neutral, citing slow agentic-AI delivery and the risk of further cloud-backlog deceleration.

Klein’s response: put AI closer to the CEO

SAP’s July 1 reorganization concentrated responsibility for AI product development. Rather than immediately replacing Muhammad Alam, the executive board member responsible for product and engineering, SAP redistributed his responsibilities after he said he would not renew his contract when it expires in March 2027. 9

Reporting on the reshuffle said Christian Klein took responsibility for Alam’s teams except Industrial AI, which moved to COO Sebastian Steinhäuser. 49 The practical message is clear: AI platform, product and execution decisions now have more direct executive attention, with less room for diffusion of accountability.

The organizational change supports SAP’s broader Autonomous Enterprise model:

• Joule is the engagement layer.
• SAP Autonomous Suite is the operational execution layer, including Industry AI.
• SAP Business AI Platform is the unifying foundation. 1

A reorganization cannot by itself create useful AI products. But it can reduce coordination friction between the teams responsible for data, governance, platform capabilities and workflow applications—the pieces that must work together for an enterprise agent to be trusted with consequential tasks.

SAP’s strongest AI asset: business context with controls

SAP’s opportunity is not to become a general-purpose frontier-model company. Its potential advantage is the business context embedded in enterprise workflows.

At Sapphire in May, SAP introduced the Business AI Platform as a unified environment combining SAP Business Technology Platform, SAP Business Data Cloud and SAP Business AI. SAP describes it as a governed foundation for building and deploying enterprise AI grounded in business context. 12

That positioning matters in processes such as closing financial books, approving procurement, managing supply-chain exceptions or administering HR activities. Useful automation in those settings needs more than a capable model. It needs to understand business entities and process states, respect permissions, maintain auditability and execute actions within defined controls.

SAP has also said its platform is intended to connect SAP and non-SAP data without forcing customers to rebuild their data estates. 3 If that works reliably in real deployments, it would make the platform more relevant to enterprises with heterogeneous technology environments.

The delivery question: announced agents versus deployed agents

SAP’s product vision is ambitious. The company has described more than 50 Joule Assistants and more than 200 specialized agents across areas including finance, spend management, supply chain, human capital management and customer experience. 1920

But the board and investors need evidence that the roadmap is becoming broadly usable software. UBS said in August that SAP had delivered 17 out-of-the-box AI agents, with another 15 in ramp-up, and viewed the company’s target of 200 agents by year-end as challenging.

The apparent contrast does not necessarily mean the product vision is invalid: vendor announcements, agents in development, agents in rollout and production deployments are different measures. It does mean headline agent counts are a poor substitute for execution evidence.

The more useful tests are whether customers can:

1. deploy agents in production with appropriate identity, access and audit controls;
2. automate multi-step work across SAP and non-SAP systems;
3. measure tangible improvements in cycle time, cost, service or risk management; and
4. expand usage in ways that strengthen cloud bookings, retention and recurring revenue.

Klein said in September that SAP was preparing 400 agents for launch that month, followed by another wave in the fall. That is a forward-looking product statement, not yet proof of sustained customer value. 3

How SAP can compete with larger platforms and faster startups

SAP faces companies with different strengths. Microsoft and Google have extensive cloud, model and developer ecosystems. Salesforce owns important customer-facing workflows. AI-native startups can often move quickly with narrower, focused products.

SAP’s plausible route is to become the trusted orchestration and execution layer for high-value business operations already running through its systems. Its core proposition is strongest where AI needs verified business data, process logic, permissions and human oversight rather than just a conversational interface.

That advantage is defensible only if SAP makes it practical. Customers will judge the company on interoperability, reliability, governance and business outcomes—not on whether it announces the largest number of agents.

What would show the strategy is working

SAP does not need to win the general-purpose AI race to establish a durable enterprise-AI position. It does need to demonstrate that its platform can make controlled automation work at scale in the environments where SAP already has deep process knowledge.

Near-term proof points include production deployments of agents in core workflows, credible controls for consequential actions, successful use across mixed data estates and evidence that AI is improving commercial momentum. The supervisory board’s reported timetable reflects the central risk: if competitors become the default AI layer above enterprise systems first, SAP’s access to valuable data and workflows may be less decisive than it appears today. 2

SAP still has a strong foundation for an enterprise-AI strategy. The question is whether its new leadership structure and platform architecture can turn that foundation into repeatable customer outcomes fast enough to meet a much less patient market.