Literal Labs, a Newcastle University spinout developing logic-based AI systems, has raised $6.2 million in pre-seed funding to bring its proprietary models to market by the end of 2025.
The pre-seed funding round is led by Northern Gritstone, a science and tech-focused investment firm from Northern England, and co-led by Mercuri, a London-based VC backing media and digital lifestyle tech. Additional investors include Sure Valley Ventures, Cambridge Future Tech SPV, and several notable angels.
At the heart of Literal Labs’ thesis is a pivot away from deep learning’s black-box complexity to a more transparent, energy-efficient form of AI: logic-based machine learning, inspired by the Tsetlin machine, a relatively underexplored architecture devised by Soviet mathematician Mikhail Tsetlin, a contemporary of AI pioneer John McCarthy.
Unlike neural networks that rely on vast layers of computation and massive datasets, the Tsetlin machine is based on propositional logic. According to recent MLPerf benchmarking, Literal’s models can deliver up to 54x faster inference speeds, while consuming 52x less energy compared to traditional deep learning techniques. The startup also claims 250x performance gains over widely used tree-based algorithms like XGBoost.
“We’re at a pivotal moment for AI adoption,” said Noel Hurley, CEO of Literal Labs and a former senior executive at Arm, where he ran the CPU division responsible for over $1 billion in annual revenue. “But current AI systems are fundamentally unsustainable. Our logic-based models offer a fresh path forward—leaner, greener, and inherently more explainable.”
Literal Labs is positioning its technology for Edge AI applications—markets where power and compute constraints are critical, such as mobile devices, IoT hardware, autonomous vehicles, and smart sensors. These environments are often underserved by traditional AI due to size and energy constraints, as well as emerging regulations requiring AI transparency and accountability.
The startup doubled its headcount to 12 in 2024, hiring Leon Fedden, former Deep Learning lead at AstraZeneca, as Chief Technology Officer. The fresh capital will fuel further expansion of the engineering team and accelerate the development of the company’s first commercial product, expected to launch later this year.
The logic-based approach may be gaining attention at a timely moment. Regulators globally—including the EU’s AI Act and proposed frameworks in the UK and US—are pressing for more auditable, explainable AI models, especially in sectors like healthcare, finance, and autonomous systems. Literal’s architecture, which produces inherently understandable decision processes, could prove a regulatory fit.
“Literal Labs is our first Newcastle University-linked investment and exemplifies the deeptech potential of the North of England,” said Duncan Johnson, CEO of Northern Gritstone. “We believe this team is on track to deliver AI technology with real societal and environmental impact.”
Esha Vatsa, Partner at Mercuri, echoed that sentiment: “Literal is redefining what efficient AI can look like. The team’s combined academic pedigree and industry leadership is a rare mix.”
The funding also highlights a broader trend in AI investing—backing alternatives to large language models and neural networks, which have recently come under scrutiny for their resource intensity and lack of transparency. While giants like OpenAI and Google AI continue to push the boundaries of scale, startups like Literal are carving out a space for targeted, sustainable AI that works outside the data center.
Read Also: Silobreaker AI Summarisation Tools to Boost Threat Intelligence Workflows
