Given Generative AI is very good at interpreting dense texts, it’s been a boon for startups attacking one of the most complex sets of texts there is: the law. We’ve thus seen an explosion of legal tech, supercharged by AI, in the last year or so.
Lawtech startup Eudia bagged $105 million only last week. Last year, London-based Genie AI raised €16 million, US-based Harvey landed a $300 million round led by Sequoia, and Lawhive raised $40 million to go after ‘main street’ US lawyers. The list goes on, but joining that list today is Luminance, which is billing itself “legal-grade” AI.
Claiming to be capable of highly accurate interrogation of all aspects of complex legal issues and contracts, Luminance raised $75 million in a Series C funding round led by Point72 Private Investments. This funding makes it one of the largest funding rounds for a pure-play legal AI company in the UK and European market. According to CrunchBase, it had previously raised $63.2 million, which means it’s raised $138 million to date.
Also participating was Forestay Capital, RPS Ventures, and Schroders Capital, as well as existing investors including March Capital, National Grid Partner,s and Slaughter and May.
Luminance was originally developed by Cambridge-based academics, Adam Guthrie (Founder and Chief Technical Architect) and Dr Graham Sills (Founder and Director of AI), and seed-funded by the late Dr Mike Lynch, founder of Autonomy, who died in a tragic accident last year.
In order to operate, Luminance uses what is called specialist AI, which it calls a “Panel of Judges.” to automate and augment a business’s approach to contracts, including generation, negotiation, and post-execution analysis.
It currently has over 700 clients spread across over 70 countries and includes companies such as AMD, Hitachi, LG Chem, SiriusXM, Rolls-Royce, and Lamborghini. Headcount has, claims the firm, tripled in North America, with the opening of three offices in San Francisco, Dalla,s and Toronto, while the company expanded its US HQ in New York.
Luminance has its own proprietary Large Language Model (LLM). Its main product, Lumi Go, lets customers send draft agreements to a counterparty and have the AI auto-negotiate on their behalf, said Luminance.
Rather than using a GPT (Generative Pre-trained Transformer) Luminance uses what it describes as an LPT (Legal Pre-trained Transformer), trained on over 150 million verified legal documents, many of them non-public, rather than open-source data, making its platform relatively defensible in a post-DeepSeek world. Other law tech startups tend to build on existing general-purpose GenAI LLMs.
Eleanor Lightbody, CEO of Luminance (who took over from the founders after the Series A round) told TechCrunch over a call:
“It’s a domain-specialized AI that being built with lawyers in mind… They need to understand that the outputs of have been validated and can be trusted, and that’s exactly what our specialized AI can achieve.”
She said the team has built its platform “understanding that different models are good at some things and other models are better at other things… What you want is to have a mixed model approach, where the models can check each other’s ‘homework’ and you can get the most accurate and the most transparent answers.”
She claimed this approach sets the company apart from others, plus, clients can “use this across the entire contract life cycle.”
In a statement, Sri Chandrasekar, Managing Partner of Point72 Private Investments, said: “We know this market well and strongly believe in the power of next-generation AI to revolutionize contracting processes across the enterprise.”
Luminance was Lynch’s third company after Darktrace and Autonomy, prior to his tragic passing.
Add comment