Kausable Raises €12M for Adaptive AI Development

German AI innovator kausable has successfully closed a €12 million Seed funding round, signaling strong investor confidence in its novel approach to artificial intelligence. The capital infusion is earmarked for advancing a groundbreaking reasoning-first AI model designed to adapt dynamically to new information without the need for extensive, costly retraining. This development addresses a critical limitation in current AI systems, which often struggle with evolving environments.

The funding round was spearheaded by prominent venture capital firms UVC Partners and Entourage. Further backing came from established German investors High-Tech Gründerfonds (HTGF) and Mätch VC. The startup also benefits from the support of influential angel investors with deep ties to leading AI research institutions and companies, including Black Forest Labs, OpenAI, Google DeepMind, Noxtua, and the European Laboratory for Learning and Intelligent Systems (ELLIS).

Johannes Haux, CEO and co-founder of kausable, highlighted the imperative for AI to mirror the world's inherent dynamism. \"The world evolves rapidly, and so should AI,\" Haux stated. \"Current AI systems are largely static and require constant, expensive retraining. We are developing a new paradigm with a world model that learns efficiently and adapts robustly on its own, unlocking significant potential in volatile sectors like robotics, energy, and finance.\" Founded in 2025 by physicists Haux, Dr. Benjamin Herdeanu (CTO), and Gregor Ramien (COO), kausable's technology stems from research connections to Heidelberg University and Black Forest Labs.

Kausable's core innovation lies in its foundational AI model, which is built on reasoning capabilities rather than brute-force data ingestion. Unlike conventional models that rely on massive datasets and frequent retraining, kausable's system is designed to learn from minimal data and adapt to unforeseen scenarios. This "world model" approach allows for the seamless integration of diverse data sources at a single comprehension layer, enabling operation in contexts that were not predictable during initial development. The company aims to deliver horizontal AI solutions capable of tackling novel problems, moving beyond the limitations of single-purpose, brittle models.

The company's recent development, TipPFN, a zero-shot forecasting model for complex systems, demonstrates this adaptive capability. TipPFN is engineered to predict rare but high-impact events, often termed "black swans," in fields such as medicine and energy. This follows a successful pre-Seed funding round of €1.5 million, underscoring the early traction and validation of kausable's vision. The underlying causal reasoning architecture of their frontier model has also been validated through collaborative research with experts from Columbia University.

Pieterjan Bouten, co-founder of Entourage, expressed enthusiasm for kausable's disruptive potential. \"Most AI models are trained to recall the past; kausable is building AI that reasons about the future,\" Bouten remarked. \"Their approach bypasses the need for ever-larger datasets and constant retraining, focusing instead on systems that infer causality and solve entirely new problems. This ambitious scientific endeavor is precisely the kind of foundational AI company we seek to support.\" The newly acquired capital will be instrumental in expanding kausable's nine-person team of elite experts and accelerating the development of its adaptive AI model.