SAP’s Billion-Euro AI Pivot Meets a Brutal Reality Check on the Stock Chart

SAP unveils Autonomous Suite with 50+ Joule AI agents, but shares drop 28% YTD. Cloud revenue target of €26B hinges on July Q2 results.

The vision is grand. The stock is anything but. When SAP took the stage at its Sapphire conference in Orlando to unveil the “Autonomous Enterprise,” the message was clear: business processes are about to be radically automated. Yet as the Walldorf-based software giant rolls out its most aggressive strategic overhaul in decades, investors have been voting with their feet. Since the start of the year, SAP’s shares have cratered by nearly 28%, a decline that no AI-powered roadmap has been able to reverse.

At the heart of the new push is the SAP Autonomous Suite, a layer that retrofits existing applications with specialized intelligent agents. More than 50 so-called Joule assistants will coordinate a network of hundreds of sub-agents, handling everything from supply-chain logistics to human resources. To give customers the ability to tailor these tools, SAP will launch Joule Studio in June, followed by an AI Agent Hub for centralized agent management in the third quarter. The concept is ambitious, but the market is demanding proof of traction.

That proof will have to come in the form of cloud revenue. SAP’s leadership is targeting cloud sales of roughly €26 billion for the full year, a figure that already includes some contribution from the AI push. The next major checkpoint is 23 July, when the company reports second-quarter results. Analysts like DZ Bank’s Armin Kremser acknowledge the strategic logic — he sees SAP on a path to becoming the central AI and data platform for enterprise software — but caution that the technology is currently an adoption theme rather than a near-term revenue driver.

To support the autonomous vision, SAP has been on a buying spree. The acquisition of data-cleansing specialist Reltio has already closed, and the purchase of data platform Dremio is expected to close in the third quarter. More eye-catching is the commitment to building proprietary foundation models. SAP is investing more than €1 billion in Prior Labs, a startup focused on tabular foundation models, and plans to establish a leading AI lab for structured enterprise data. Its model, SAP-RPT-1.5, is designed to outperform large language models on the kind of structured, numeric data that is the lifeblood of corporate systems. A separate €60 million investment in Berlin-based automation startup n8n valued that company at $5.2 billion.

The data infrastructure is further strengthened by deep integrations with cloud hyperscalers. Anthropic’s Claude will serve as the primary engine for Joule agent logic in HR, procurement and supply chains. Amazon Web Services is integrating SAP Business Data Cloud with Amazon Athena via zero-copy connectivity, while Google Cloud and Microsoft are enabling agent interoperability across external frameworks. These partnerships, combined with a partner fund running into the hundreds of millions, are meant to ensure the agents can actually act on real-world data — not just deliver polished demonstrations.

Yet none of this has lifted the stock out of its technical slump. On Friday, SAP shares recovered to €145.84, a gain of 3.24% on the session, but that came just days after the stock touched a new 52-week low of €137.62. The price remains roughly 25% below its 200-day moving average, a classic signal of a persistent downtrend. From its January levels, the stock is still down 27.80%.

The disconnect between SAP’s product narrative and its share price is a subplot that will define the rest of 2026. The company has delivered a broad bundle of tools: a unified AI platform, a data layer, specialized agents, deep cloud partnerships and migration aids that promise to reduce workload by more than 35%. The market, however, is waiting for those pieces to convert into measurable cloud bookings. With the next product milestones scheduled for the summer and the €26 billion cloud-revenue target on the table, the burden of proof now rests squarely on execution.