SAP has announced plans to acquire German AI startup Prior Labs and invest €1 billion (approximately $1.16 billion) into the company over the next four years, signaling a major push to strengthen its enterprise AI capabilities. The move comes as SAP intensifies efforts to defend and expand its position in an increasingly AI-driven enterprise software market.
Founded just 18 months ago, Prior Labs specializes in “Tabular Foundation Models” (TFMs), AI systems designed specifically to work with structured enterprise data stored in tables and databases. Unlike traditional large language models that primarily process text, TFMs are optimized for business-critical datasets such as finance, procurement, HR, and operational systems—areas central to SAP’s ecosystem.
As part of the acquisition, SAP plans to turn Prior Labs into a dedicated AI lab focused on structured data intelligence. Reports indicate that the deal represents one of Germany’s largest recent AI startup exits, with the founders reportedly receiving a substantial cash payout.
At the same time, SAP is tightening control over how AI agents interact with its software ecosystem. The company has updated its API policies to restrict access primarily to “SAP-endorsed architectures,” including selected technologies such as Nvidia’s NemoClaw framework. This approach reflects SAP’s preference for a tightly governed AI environment rather than allowing unrestricted third-party agent integration.
The strategy stands in contrast to competitors like Salesforce, which are taking a more open approach toward enterprise AI agents and integrations. SAP’s position highlights a growing divide in enterprise AI strategies, with some companies prioritizing ecosystem control and governance while others focus on flexibility and interoperability.
The investment also underscores the broader shift toward specialized enterprise AI infrastructure. As businesses increasingly demand AI systems capable of working with complex operational data, companies are moving beyond generic language models toward domain-specific AI architectures built for enterprise workflows and decision-making.
 into the company over){kind=link}