Recursive Superintelligence secures £483m to automate scientific discovery with self‑improving AI
This article covers Recursive Superintelligence, a London AI startup, which has closed a growth funding round of about £483.2m ($650m) and emerged from stealth with a valuation of roughly £3.5bn. The funding will support development of automated scientific discovery and systems that can safely improve themselves, and is relevant to advanced AI research, investors and the UK startup ecosystem.
Recursive Superintelligence, a London AI startup, has closed a growth funding round of about £483.2m ($650m), valuing the company at roughly £3.5bn as it emerges from stealth. Founded in 2025 by a group of former research leaders from OpenAI, Google DeepMind, Meta AI, Salesforce AI and Uber AI, the company says it will focus on automated scientific discovery and systems that can safely improve themselves.
Why it matters
A near half‑billion‑pound debut round for a company less than two years old underlines how much capital is still chasing advanced AI research and talent. Recursive’s backers include strategic names from both venture capital and chip makers, signalling investor appetite for projects that combine large‑scale compute, novel algorithms and safety work. The firm’s stated goal — building processes that automate scientific discovery and enable recursive self‑improvement — touches on long‑standing technical and governance debates about capability growth and safe deployment.
Inside the product
Recursive says it is developing an “open‑ended” approach to machine learning: systems that generate and test their own hypotheses, environments and learning challenges. The announcement lists a wide set of research strands the team claims experience in, including open‑ended algorithms, quality‑diversity methods, AI‑generating algorithms, self‑improving coding agents, automated red teaming, capability discovery, prompt engineering, retrieval‑augmented generation and foundational world models.
The company describes this stack as a way to automate parts of the scientific method — producing experiments, iterating on failures and scaling discoveries without relying solely on hand‑designed methods. It currently employs more than 25 people and maintains offices in London and San Francisco.
Who are the investors?
The round was led by GV (formerly Google Ventures) and Greycroft, with major participation from AMD Ventures and NVIDIA. The mix of traditional venture backers and chip‑maker investors suggests funding is intended to support both algorithmic development and substantial compute needs.
GV is a long‑standing backer of ambitious AI projects and provides access to networks inside Alphabet; Greycroft is a US venture firm with a broad portfolio across enterprise and consumer tech. AMD Ventures and NVIDIA bring direct hardware alignment: both firms invest in startups that will drive demand for specialised silicon and GPUs. The presence of these participants indicates investor confidence in Recursive’s ability to bridge research and deployable compute‑intensive systems.
If you're researching potential backers in this space:
- Explore our list of AI venture capital firms in the UK
- Browse our directory of AI angel investors in the UK
Founder perspective
Recursive’s public statements emphasise both ambition and safety. The company framed its approach as a continuation of open‑ended evolutionary processes that produced human intelligence, arguing that similar mechanisms could accelerate scientific discovery when automated.
This will likely be the fastest path to superintelligence
The excitement for what we are building has been incredible. The playbook we create will soon allow us to revolutionise every scientific discipline
The potential benefits for humanity of safely creating such an advance cannot be overstated
Throughout, we will prioritise safety. We must make sure the system helps humanity flourish by maximising the benefits while reducing risks
Those quotes were released by the company in its announcement; no individual spokesperson was named. The founders’ previous roles at major AI labs give the team notable technical credibility, but the combination of scale, secrecy and lofty aims will draw scrutiny from researchers, policymakers and other AI investors.
The bigger picture
This funding round comes as governments and regulators in the UK and EU increase attention on advanced AI systems, including labelling, testing and risk mitigation. Large sums going into early‑stage research companies complicate the policy conversation: more private capital accelerates capability development but also raises questions about transparency, third‑party evaluation and public oversight.
For the UK ecosystem, Recursive’s London base and international investor list are a reminder that cutting‑edge AI projects can be anchored in the UK while drawing global capital and talent. The deal also highlights how closely intertwined algorithmic research and hardware supply are becoming, a factor that will shape competition among startups and incumbent cloud and chip providers across Europe and beyond.
The announcement is a sign of continued investor willingness to back ambitious AI research in the region, even as debates over safety, compute concentration and governance intensify.
They've invested in Recursive Superintelligence
Get to know these AI investors
Click here for a full list of 7,589+ startup investors in the UK