Former DeepMind staff raise $50M for science-focused AI lab Inherent

London lab Inherent officially emerged from stealth on Wednesday with a $50 million funding round. It is one of the largest seed rounds for AI startups in Europe in 2026. The news suggests that European researchers and investors can compete with American giants in artificial intelligence, building companies that can attract world-class capital.

A team from DeepMind, Microsoft and Reka

Inherent was founded by researchers from three of the world's leading AI organizations: DeepMind, Microsoft and Reka. This is not just another startup but a project built by people who know from the inside how modern science around large language models is done, from first principles to scaling systems.

Such a team is already a sign the project is serious. These people understand how cutting-edge AI systems are developed, what mistakes newcomers often make and where the greatest potential for breakthrough innovation lies.

The investor syndicate includes:

• Index Ventures and Radical Ventures as co-leads of the round
• Nvidia's venture arm NVentures, the voice of the most powerful maker of AI chips
• Ex/Ante, Metaplanet, Macroscopic Ventures and Mythos Ventures

The investor lineup indicates that serious money sees Inherent as a company with huge potential and a unique approach to the problem its founders have set out to solve.

The mission: teaching AI to pick worthwhile questions

Based on press headlines and the company's name, Inherent is working on AI that helps scientists determine which scientific questions are truly worth investigating. At first glance this sounds abstract, but it is one of the most ambitious ideas in modern fundamental science.

The problem is that scientific progress often stalls not for lack of hypotheses but for lack of good questions. Researchers can spend years on a problem that turns out to be a dead end, or pass by a direction that would have radically changed an entire field. A tool that can use AI to predict how promising a scientific question is could dramatically change the pace of fundamental research. It would cut the resources wasted on unpromising directions and let the best minds focus on what really matters.

Europe's challenge to the American AI ecosystem

$50 million is a significant round even in 2026, when training large models costs billions. But for a seed round it is an impressive sum, allowing a European company to hire top researchers and buy the computing power it needs at a level that competes with its American counterparts.

Inherent's emergence as one of the largest European AI startups of 2026 shows that investors believe in Europe's talent pool. Not everything is decided in San Francisco alone. London, Zurich, Berlin and other hubs can produce world-class researchers able to take on any competition.

What this means

Fundamental science is starting to actively seek AI-based support tools to speed up its progress. This could change both the pace of scientific development and its main directions as a whole.

If Inherent can prove that AI can help scientists choose the right questions, it could become a turning point, when artificial intelligence stops being merely a tool and becomes a strategic partner in fundamental scientific discovery.