SAP Bets $1.16 Billion on Prior Labs to Build European AI Frontier Lab

--- headline: "SAP Bets $1.16 Billion on Prior Labs to Build European AI Frontier Lab" slug: sap-prior-labs-1b-acquisition category: business story_number: "02" date: 2026-05-24 sources: - https://news.sap.com/2026/05/sap-to-acquire-prior-labs-establish-frontier-ai-lab-europe/ - https://techcrunch.com/2026/05/05/sap-bets-1-16b-on-18-month-old-german-ai-lab-and-says-yes-to-nemoclaw/ - https://sifted.eu/articles/sap-acquire-german-ai-startup-prior-labs - https://priorlabs.ai/blog-posts/priorlabs-next-chapter ---

# SAP Bets $1.16 Billion on Prior Labs to Build European AI Frontier Lab

In a deal that could redraw the map of enterprise AI, SAP has agreed to acquire Prior Labs — an 18-month-old startup out of Freiburg, Germany — and pledged to invest more than €1 billion (approximately $1.16 billion) over four years to grow it into what the German software giant calls a "globally leading frontier AI lab" for structured data.

The acquisition, announced May 4 and pending regulatory approval expected in Q2 or Q3 of 2026, is a striking bet on a niche but strategically critical corner of AI: Tabular Foundation Models, or TFMs. Where large language models parse prose, TFMs are built to reason natively over the rows and columns of spreadsheets, ERP exports, and business databases — exactly the data that underpins SAP's sprawling enterprise software portfolio. The deal terms were not disclosed, though sources cited by Pathfounders describe it as an almost entirely cash transaction, with well over half a billion dollars going to the founders up front.

A Tiny Lab, A Very Large Check

Prior Labs was founded in 2024 by Frank Hutter, Noah Hollmann, and Sauraj Gambhir. Before this deal, it had raised a single €9 million pre-seed round, led by Balderton Capital in February 2025. The acquisition is widely described as one of the largest-ever venture outcomes for a German startup, made all the more extraordinary by its speed: the company had barely cleared its first birthday before landing a ten-figure commitment.

"Over the last 18 months, Prior Labs has built an incredible team, increasing the velocity in tabular foundation models," CEO Frank Hutter said in the announcement. "Joining the SAP family gives us the resources, data environment and customer reach to take this category to its full potential."

The core technology — TabPFN — has earned genuine academic credibility. Published in Nature and cited over 1,000 times, the model topped the TabArena benchmark with its latest version, TabPFN-2.6, matching the accuracy of a four-hour automated machine learning pipeline in a single inference call. The open-source releases have accumulated more than three million downloads, building a developer ecosystem that SAP has explicitly committed to sustaining.

Why SAP Needed This, Now

SAP is navigating a difficult moment. Its stock has slid in 2026 amid what analysts are calling the "SaaSpocalypse" — a structural threat to traditional SaaS incumbents from AI agents capable of automating workflows that once required enterprise software licenses. Against that backdrop, the company has been making a series of calculated moves: it recently completed the acquisition of Reltio, announced a deal to buy Dremio for data unification, and had previously invested in Anthropic, Cohere, and Aleph Alpha.

The Prior Labs deal is different in kind, not just degree. Rather than buying distribution or an adjacent SaaS product, SAP is acquiring fundamental research capability and attempting to own a new AI category from its scientific foundations.

"Early on, SAP recognized that the greatest untapped opportunity in enterprise AI wasn't large language models; it was AI built for the structured data that runs the world's businesses," SAP CTO Philipp Herzig said. "We built SAP-RPT-1 to prove that conviction for enterprise data. Prior Labs has built a leading TFM on public benchmarks and built one of the leading research teams in this category. Combining their frontier model work with enterprise data and customer reach is how we intend to lead this category globally."

SAP-RPT-1, the company's own relational pretrained transformer, was an earlier signal of this conviction — Prior Labs is not a detour but a shortcut to a destination SAP was already heading toward.

The Independence Question

The structure of the deal is notable. Prior Labs will continue operating as a legally independent entity, retaining its brand, Freiburg headquarters, Berlin and New York offices, and its open-source commitments. SAP's investment flows through the lab, not over it. The research team — recruited from Google, Apple, Amazon, Microsoft, Jane Street, Goldman Sachs, and CERN — stays intact, and the lab will continue publishing academic work.

Scientific advisory heft is not thin: Yann LeCun, ACM Turing Award winner and founder of AMI Labs, and Bernhard Schoelkopf, director of the Max Planck Institute for Intelligent Systems and ELLIS president, will both serve on Prior Labs' scientific advisory board. The founders say the mission is unchanged; what changes is the resource envelope and the path to deployment.

"What changes is the resource envelope and the deployment reach. We get to work on research problems that previously sat outside what was feasible," the founding team wrote in a blog post announcing the deal.

The commercialization runway is significant. Prior Labs' models will integrate with SAP AI Core, SAP Business Data Cloud, and the company's Joule agentic layer — giving the startup direct access to SAP's installed base across finance, procurement, HR, supply chain, and manufacturing. Investors including Atlantic Labs and XTX Ventures, alongside angels such as Black Forest Labs co-founder Robin Rombach and Hugging Face co-founder Thomas Wolf, all receive an exit from a company they backed just over a year ago.

Analysis: A Two-Front War

The acquisition signals that SAP is fighting on two fronts simultaneously. On one front, it needs to make its own software smarter, faster than its customers lose confidence in traditional SaaS models. On the other, it is actively managing which external AI agents can reach its systems.

In the same week as the Prior Labs announcement, SAP clarified its API policy to effectively block unauthorized AI agents — including OpenClaw and similar frameworks — from accessing SAP products. Only SAP-endorsed architectures are permitted, which currently include SAP's own Joule Agents (still in beta) and, by recent arrangement with Nvidia, the NemoClaw platform. This is a deliberately closed-garden posture, contrasting sharply with Salesforce's Headless 360 strategy, which allows enterprise customers to plug in any agent framework they choose.

The tension in this approach is real. SAP's enterprise customers increasingly want flexibility in their AI tooling, and locking the agent gateway while building frontier capabilities in-house is a bet that SAP can out-innovate the open alternatives before the walled garden becomes a competitive liability.

Balderton Capital partner James Wise, whose firm backed Prior Labs from the start, captured the underlying thesis in a LinkedIn post on the deal: "Most of the world's valuable data still sits in spreadsheets and databases. Replacing hand-tuned gradient-boosted trees with a one-shot model will do for structured data what LLMs have done for unstructured." He also called the outcome "one of Germany's biggest ever venture outcomes" — arriving with unusual speed.

What Comes Next

The deal is expected to close in Q2 or Q3 of 2026. Once integrated, SAP plans to use Prior Labs' capabilities to let business users run natural-language queries, generate predictions, and run "what-if" scenarios on structured enterprise data without requiring data science expertise — essentially bringing the TabPFN interface to SAP's hundreds of thousands of enterprise customers worldwide.

TabPFN-3, the next generation of the model, is described as imminent. The goal, as the founders frame it, is to push beyond correlation into causation: models that do not merely predict what will happen, but explain why.

For Europe's AI ecosystem, the symbolism is hard to ignore. A frontier research lab, funded at scale, staying on European soil, committed to open research — it is precisely the kind of outcome European tech policy has been arguing is achievable. Whether Prior Labs can preserve its scientific velocity inside one of the world's largest enterprise software companies is the question that will take years to answer.

For now, the numbers tell their own story: an 18-month-old startup, a single seed round of €9 million, and a commitment north of a billion euros. In the compressed timeline of the current AI race, that math is its own kind of benchmark result.

---

Sources: SAP News Center, TechCrunch (Anna Heim), Sifted (Anne Sraders), Prior Labs blog