London has evolved into the undisputed command centre for European artificial intelligence. It is no longer just a research hub; it is the operational headquarters for the world’s most ambitious engineers. The ecosystem here has progressed beyond the experimental phase into industrial-scale deployment, resulting in a density of deep tech talent unmatched anywhere else on the continent. This is where Google DeepMind started, and it is where the next generation of sovereign intelligence is being forged.
This is the exclusive list of founders who are building the engines of the future. These companies are teaching cars to drive using intuition, designing life-saving drugs with algorithms, and creating digital workers that never stop working.
Here are the 10 AI startups in London you need to know in 2026.
Wayve
Alex Kendall founded Wayve to destroy the old map-based approach to self-driving cars. They are building Embodied AI that allows vehicles to learn from experience rather than rigid rules. The company raised over $1 billion in a Series C round led by SoftBank to unleash this technology on public roads. Their end-to-end deep learning models enable cars to navigate complex urban environments in London just by seeing and processing the world like a human driver.
Synthesia
Victor Riparbelli and Steffen Tjerrild built Synthesia to eliminate the need for cameras and studios. Their platform allows anyone to create professional video content using AI avatars that speak over 120 languages perfectly. Valued at $1 billion, they have become the go-to video production suite for Fortune 500 companies. They have recently unveiled expressive avatars that can convey human emotions, effectively transforming video production into a software-driven process.
Isomorphic Labs
Demis Hassabis launched Isomorphic Labs with a singular mission to reimagine drug discovery from first principles. Operating as a standalone commercial venture under Alphabet, they use the AlphaFold technology to model life at a molecular level. They recently signed multi-billion-dollar partnerships with Eli Lilly and Novartis to design novel therapeutics. This is not just a startup; it is a digital biology factory that aims to cure diseases by treating biology as an information processing problem.
11x
Hasan Sukkar created 11x to build the automated workforce of tomorrow. Their flagship digital worker, Alice, is an AI sales development representative that autonomously sources leads, writes emails, and books meetings. Backed by Benchmark and Y Combinator, they are aggressively scaling their fleet of digital employees. Their systems don’t just assist humans; they replace entire workflows, allowing revenue teams to operate on autopilot 24/7.
PolyAI
Nikola Mrkšić founded PolyAI to fix the broken state of customer service voice assistants. Their voice AI sounds indistinguishable from that of a human and can handle complex conversations across various industries. Having raised $50 million, they serve global brands like FedEx and Marriott. Their models are pre-trained on billions of conversations, enabling them to understand accents, interruptions, and digressions more effectively than most human agents.
AutogenAI
Sean Williams started AutogenAI to automate the high-stakes world of bid writing. Their platform utilises large language models to enable companies to write winning tender proposals in minutes, rather than weeks. They became one of the fastest-growing SaaS companies in the UK, securing substantial funding to expand into the US market. By synthesizing corporate knowledge into persuasive prose, they help clients win government contracts worth millions.
Unitary
Sasha Haco and James Thewlis founded Unitary to make the internet safe without suppressing free speech. Their multimodal AI analyses video content in context to detect harmful material like hate speech or sexual abuse. Unlike simple keyword filters, their system understands visual and audio nuance. Backed by Index Ventures, they process billions of video frames daily for major social platforms, acting as the intelligent immune system for the creator economy.
Robin AI
Richard Stumpf built Robin AI to give every lawyer a brilliant copilot. Their legal assistant reviews contracts, suggests edits, and negotiates terms on their behalf. As an early partner of Anthropic, they leverage some of the most advanced models in the world to understand legal semantics. They recently raised $26 million to expand into the US legal market, helping legal teams reduce contract review time by 85%.
LabGenius
James Field founded LabGenius to engineer new proteins using a closed loop of AI and robotics. Their EVA platform uses machine learning to design antibodies and then physically builds and tests them in a robotic lab. The data from these experiments feeds back into the model, making it smarter with every iteration. This evolutionary engine enables them to discover potent therapeutic antibodies that human scientists would never have conceived.
Haiper
Yishu Miao and Ziyu Wang, both alumni of Google DeepMind, founded Haiper to build the next generation of perceptual video foundation models. Their technology enables users to create high-quality video content from simple text prompts. Emerging from stealth with nearly $14 million in seed funding, they are challenging the dominance of US video models. Their focus is on building a powerful creative tool that gives creators total control over video composition and motion.