Founded in 2016, TechRound is the UK’s leading online publication and news platform dedicated to startups, entrepreneurs, and the wider technology ecosystem. It placed Conscium second in its latest roundup of the UK’s top AI startups.
“The UK has built so many AI companies working in all industries and data from Beauhurst shows that many of the best funded names have raised hundreds of millions of pounds, with backers spanning global venture capital groups and public bodies. This flow of money has helped products reach real customers at speed.
Healthcare is one of the most active areas. Exscientia, founded in 2012 in Tayside, uses AI to test millions of drug molecules and predict which ones could work best. Beauhurst says the company raised £299m before its stock market listing, alongside grants from Horizon 2020 and the Bill and Melinda Gates Foundation.
Isomorphic Labs also works on drug discovery. It raised £464m in a single round from Alphabet, Google Ventures and Thrive Capital. The funding supports AI research designed to speed up the hunt for new treatments.
Which UK AI Companies Have Attracted The Most Money?
Transport and chips have also gotten some heavy backing. Wayve raised £822m in February 2024 from Nvidia, SoftBank and M12, reaching a post money valuation of £2.22b, according to Beauhurst. Graphcore raised £527m before SoftBank acquired the business for £462m, keeping its chip technology active under the same name.
Cybersecurity group Quantexa raised £421m and works with banks and public bodies to flag illegal activity. Health service provider Cera raised £479m to build data driven homecare and telehealth services through a mobile app.
Here are the 10 UK AI companies with the highest total capital raised, based on Beauhurst data:
- Wayve £1.01b
- OneTrust £699m
- Graphcore £527m
- Cera £479m
- Isomorphic Labs £464m
- Quantexa £421m
- Thought Machine £397m
- Lighthouse £378m
- Multiverse £318m
- Exscientia £299m
These businesses cover autonomous driving, privacy software, chips, banking platforms, travel pricing and workplace training. Most operate from London, with activity also in Scotland and the south west.
How Do Big AI Firms Influence Smaller UK Companies?
Large AI companies influence daily work for smaller firms through software tools and services. The Mole Valley Chamber report says UK SMEs account for about 99% of all businesses. Between 35% and 39% already use AI powered tools, up from 25% in 2024. Another 31% are considering adoption, taking total engagement close to 70%.
The same report says government reviews estimate that safe use of AI could add £47 billion to the economy over ten years, with productivity growing by 1.5% each year. Within SMEs, 45% say AI speeds up routine tasks and 39% say it lowers staff workload.
Finance software shows strong uptake. Sage Copilot offers “smart nudges ahead of VAT deadlines,” according to the study. Xero’s AI can match up to 90% of bank lines automatically, with error reduction of 37%.
This connection between large AI builders and smaller firms continues to deepen as more tools reach the market.
Top UK AI Startups:
Parisi
According to research, more than 50% of an organisation’s value is driven by perception and reputation, yet most communications teams lack a real-time, evidence-based view of how their activity is shaping stakeholder opinion and business outcomes. Parisi sought to address this with Parisi:Blue.
As a first mover in applying AI to corporate communications, we developed an AI-powered intelligence platform that gives organisations a real-time view of how their communications shape perception, influence stakeholders and drive business outcomes.
Parisi:Blue combines global media and parliamentary data with custom data sources to turn complex information into clear, actionable insight. It brings together performance metrics – reach, sentiment, importance and relevance – with contextual data showing who is driving the conversation and where it is happening, helping comms teams identify key voices, spot emerging trends and focus engagement where it matters most.
The platform integrates additional data sources, including market trends, economic indicators and website traffic, providing a comprehensive picture of the factors shaping influence and opportunity in each sector.
Unlike traditional monitoring tools, it goes beyond simple mention counts. At the heart of the system sits the Parisi:Blue Score™, a proprietary metric that brings together volume, sentiment, importance and relevance into a single number, so teams can track and understand changes in communications performance over time.
Parisi:Blue is the only platform to integrate print, online, broadcast and parliamentary data in one system alongside clients’ own datasets. By providing a live view, Parisi:Blue enables organisations to act early, manage risk effectively, and measure the return on investment and effectiveness of their communications in building resilient reputations in an increasingly complex environment.
Conscium
Conscium is a UK-based AI safety company founded in 2024 by Calum Chace, AI author and academic, and Daniel Hulme, one of the UK’s most successful AI entrepreneurs (Satalia and Faculty). Conscium’s mission is to develop safe and efficient AI that aligns with ethical standards and human values. It has recruited some of the world’s leading neuroscientists and computer scientists to its advisory board, including Karl Friston, Mark Solms, Anil Seth and Megan Peters.
Conscium has developed a platform to verify AI agents for accuracy, responsiveness, and other performance indicators, such as fairness, explainability, and alignment. The technology will test AI agents deployed across industries like marketing, financial services, consulting among others, as they continue to deploy AI agents at scale. Gartner predicts that 40% of apps in use by enterprises by the end of this year will incorporate task-specific AI agents, up from just 5% in 2025. That means there will be a huge number of agents to put through their paces.
Conscium stands as a shining example of UK AI innovation at a time when business leaders are now asking harder questions about trust, control and verification.
PEAK:AIO
PEAK:AIO is a UK-based leader in AI infrastructure, specialsing in high-performance storage for AI and HPC. The team behind PEAK:AIO led the development of software-defined storage in 2000 and developed industry-leading frameworks that are still used and licensed in enterprise storage today. Recognizing the differences in AI workflows and data requirements, PEAK:AIO was built from the ground up to match the exact needs of AI rather than repurposing existing technologies.
Deployed on server hardware from vendors such as Dell, PEAK:AIO creates AI Data Servers that deliver ultra-low latency and high bandwidth, optimising GPU utilisation. The platform is ideally suited to mid-scale GPU clusters where traditional storage solutions are often complex and expensive. It includes enhanced performance RAID protection, world-leading kernel-compliant connectivity options and protocols that deliver high performance while remaining simple to deploy and manage.
PEAK:AIO allows organisations to start as small as 30TB and grow seamlessly up to 360TB per 2U as AI projects scale, eliminating the need for upfront overinvestment. Offering the first fully Open pNFS platform for AI and HPC, PEAK:AIO lets teams expand from single nodes to superclusters.
Through its OEM partnership with PNY, PEAK:AIO is established, proven and at the heart of some of the world’s leading AI projects, providing a flexible, cost-effective and high-performance foundation for modern AI applications.
Phoebe
Phoebe.ai is a London-based startup building the immune system for software: AI agents that help developers by continuously detecting, diagnosing and preventing software failures.
As AI accelerates the creation of new software, complex systems are constantly changing and more opaque than ever. Debugging has remained a stubbornly manual activity. Phoebe helps automate the process, reducing the time to resolve alerts and incidents by up to 90% and mitigating problems before they cause disruption.
The company was founded in 2024 by long-time collaborators Matt Henderson and James Summerfield. The pair have now worked together across four companies, including their first startup Rangespan, a retail analytics company acquired by Google in 2014. Following the acquisition, both worked at Google, and later reunited again at Stripe, where James led software engineering teams and Matt was CEO of Stripe Europe.
Phoebe grew out of a shared frustration Matt and James experienced across organisations – where the most experienced engineers would frequently be pulled away from building new features and instead into reactive work, troubleshooting problems to keep systems running.
Phoebe serves companies across fintech, digital commerce and software-as-a-service, helping them improve reliability, reduce operational risk and free engineers from time-intensive manual investigation work. The company raised $17 million in seed funding in 2025, with the round led by GV and Cherry Ventures.
Skaiy
Skaiy is a deep-tech solar intelligence company building next-generation forecasting infrastructure for renewable energy systems. Co-founder and inventor Dr Tasmiat Rahman has researched Solar PV for 15 years, including through a PHD at Imperial College and is now a research fellow at Southampton University where Skaiy’s technology has been developed. His other co-founder and CEO is serial green energy entrepreneur Duncan Grierson. The company has been supported by the University of Southampton’s accelerator, Future Worlds.
Skaiy was founded to commercialise this research and address one of the most pressing challenges facing modern electricity grids: volatility driven by rapid growth in solar power. The company delivers ultra-accurate, site-specific solar power forecasts using proprietary sky cameras, optical sensors, and AI-driven computer vision models deployed directly at generation sites. By providing real-time, hyper-local “nowcasts,” Skaiy enables solar operators, battery storage providers, energy traders, and grid operators to reduce imbalance penalties, curtailment, and operational inefficiencies.
Since incorporating in 2025, Skaiy has progressed from publicly funded research to live deployment at the Chilbolton Observatory, supported by UK Research and Innovation. The technology demonstrates significantly improved accuracy compared with traditional satellite-based forecasts. The company is now preparing for industrial pilots and commercial rollout across the UK and Europe.
Rahman, has recently been appointed to the Royal Society of Engineering Enterprise Fellowship, recognising his significant contributions to solar energy forecasting, engineering research, and innovation in the UK’s clean energy sector.
More from Artificial Intelligence
- Amazon To Cut 30,000 Jobs: What Does This Say About the Future of Tech Work?
- What Is Physical AI? The Buzzword That Dominated CES 2026
- UCL Founded AI Startup Becomes First UK Tech Unicorn Of 2026
- Before AI Comes Clean Data
- Cyber Criminals Are Using LLMs For Phishing Attacks, Here’s How
- China’s AI Stocks Are Rallying Again – And This Time, It’s About Infrastructure, Not Hype
- 76% of Americans Plan On Learning New AI Skills in 2026
- How Are AI Systems Affecting Cloud Hosting Costs?
CultureAI
CultureAI is the AI Usage Control platform built for the reality of modern work: employees are already using AI tools—often without approval—and sensitive data is flowing into models every day. Blocking AI doesn’t work. Logging it after the fact doesn’t help. CultureAI exists to enable safe, compliant, measurable AI usage without slowing teams down or killing innovation.
Where legacy tools fall short, CultureAI operates at the point of real risk: human interaction with AI inside browsers and SaaS applications. It provides clear visibility into how AI is actually used across sanctioned tools, shadow AI, and embedded AI features that traditional controls miss entirely. At the prompt level, CultureAI detects risky behaviour, understands context and intent, and intervenes in real time—coaching users before mistakes turn into incidents.
Unlike traditional DLPs or CASBs that rely on static rules and blunt controls, CultureAI manages AI use like a workflow, not a threat. Adaptive, role-aware policies allow organisations to guide behaviour with soft nudges, enforce guardrails when necessary, and apply controls that flex with real-world usage. Privacy-first by design, CultureAI avoids surveillance while delivering audit-ready reporting for security, compliance, and leadership teams.
Deployed easily via a lightweight browser extension, CultureAI gives organisations immediate insight into AI usage patterns, risk hotspots, and adoption trends—often within hours. The result is a clear move to the top-right quadrant: high AI adoption paired with high security and strong governance.
CultureAI empowers teams to use AI without fear—unlocking productivity, protecting sensitive data, and enabling organisations to innovate faster, with confidence.”
