SAP doubles down on AI with dual acquisitions of Prior Labs and Dremio
German software giant SAP has struck separate deals to acquire Prior Labs, a specialist in tabular AI models, and Dremio, a data lakehouse platform provider, as part of a broader strategy to embed artificial intelligence deeper into its enterprise software stack.
Financial terms for both transactions have not been disclosed. However, SAP has confirmed it will invest more than €1 billion ($1.17 billion) over four years to build a frontier AI lab in Europe, with Prior Labs operating as an independent entity within the company. The Prior Labs deal is expected to close in the second or third quarter of 2026, pending regulatory approvals.
Prior Labs is best known for its Tabular Foundation Models (TFMs), which are designed to make predictions on structured business data — a domain where traditional large language models often fall short. SAP plans to integrate these capabilities into its existing AI infrastructure, including SAP AI Core and SAP Business Data Cloud, to allow business users to run predictive analytics using natural language prompts without needing deep technical expertise.
“Prior Labs has built a leading TFM on public benchmarks and assembled one of the top research teams in this category,” said Philipp Herzig, SAP’s chief technology officer. “Combining their frontier model work with enterprise data and customer reach is how we intend to lead this category globally.”
The Prior Labs team includes prominent AI researchers, with Yann LeCun and Bernhard Schoelkopf joining the scientific advisory board. Its open-source tool TabPFN has already been downloaded more than three million times, and SAP has committed to keeping the project open source. The latest model, TabPFN-2.6, currently leads benchmark performance for TFMs, offering instant predictions on structured data without the complexity of traditional machine learning pipelines.
Meanwhile, the acquisition of Dremio is aimed at solving a persistent headache for enterprise AI projects: fragmented and siloed data. Dremio’s data lakehouse platform supports open formats like Apache Iceberg and Apache Arrow, eliminating the need for data conversion or relocation. SAP says the integration will streamline enterprise analytics and make SAP Business Data Cloud more compatible with both SAP and non-SAP data sources.
“Enterprise data is often scattered and lacks context, which slows down AI initiatives,” a SAP spokesperson noted. “Dremio’s serverless architecture and open catalog approach will help unify data access across systems.”
SAP plans to introduce a unified, open catalog using Apache Polaris and the Apache Iceberg REST API, enabling seamless data discovery and semantic management across the enterprise. The Dremio acquisition is expected to close in the third quarter of this year, subject to regulatory clearance.
Industry reactions have been mixed. Dr. Elena Marchetti, an enterprise AI strategist at a London-based consultancy, called the move “a smart, long-term bet.” She said: “SAP is clearly trying to own the data layer and the AI layer simultaneously. That’s ambitious, but if they pull it off, it could redefine how enterprises use structured data.”
Tomás Reyes, a data engineer at a mid-sized German manufacturing firm, was more skeptical. “SAP has a history of buying promising tech and then strangling it inside their ecosystem. I hope Prior Labs stays independent enough to keep innovating. Otherwise, this is just another acquisition that looks good on a press release.”
Linda Chu, a product manager at a SaaS startup, was blunt: “Honestly? I’m tired of these big ERP vendors pretending they’re suddenly AI companies. They’re just buying their way in. Let’s see if they can actually ship something useful instead of just hoarding talent.”
The dual acquisitions underscore SAP’s determination to compete in the AI arms race, particularly against cloud rivals like Microsoft and Google. By combining frontier AI research with a unified data platform, SAP is betting that the future of enterprise software lies in making AI accessible to non-technical users — without sacrificing the depth and specificity of business data.
This article was originally published by Verdict, a GlobalData owned brand. The information on this site has been included in good faith for general informational purposes only. It is not intended to amount to advice on which you should rely, and we give no representation, warranty or guarantee, whether express or implied as to its accuracy or completeness. You must obtain professional or specialist advice before taking, or refraining from, any action on the basis of the content on our site.