Artificial intelligence is beginning to transform professional sectors that have long remained on the fringes of advanced digitalization. After marketing functions, software development or customer support, it is now the intellectual property market which is entering a deeper automation phase. In Pau, the deeptech startup LIPSTIP announces the launch of a specialized AI platform for intellectual property professionals, with a clear ambition: to accelerate and make legal analysis linked to brand protection more reliable.
Founded in 2023 by Nicolas Girardin, Alexander Lerbs and Nicolas Vonthron, LIPSTIP develops an artificial intelligence infrastructure specifically designed for industrial property consultancies. Unlike the general models used in many emerging legal tools, lipstip claims an architecture trained on issues specific to trademark law and case law analysis. The platform is already deployed in six European firms.
The technological heart of the solution is based on the automation of the analyzes necessary for opposition and trademark protection procedures. The platform compares products and services, analyzes similarities between distinctive signs and assesses the risk of confusion according to several dimensions: phonetic, visual, semantic and conceptual. The whole is based on a base of case law covering several hundred thousand decisions.
This approach responds to a structural limitation of the sector; in many specialized firms, a significant part of the work still relies on lengthy manual research. Opposition procedures often require several hours of documentary analysis in order to identify comparable decisions, assess their relevance and then extract legally usable elements. lipstip seeks precisely to reduce this operational burden.
The startup claims that its platform allows analysis a hundred times faster and detection accuracy twice as high as general AI. Beyond the claimed technical performance, the company’s positioning above all reflects a broader evolution of the AI market: the rise of specialized models capable of reproducing business reasoning rather than simple conversational capabilities.
In regulated professions, this distinction becomes central. General models operate essentially by statistical probability, with margins of error that are difficult to accept in legal contexts. lipstip instead attempts to build an AI capable of reproducing analysis logic close to that used by intellectual property experts.
“IP professionals are the only ones who can reduce the protection gap for European companies, but they have been held back by tools that have not evolved with them,” says Nicolas Vonthron, co-founder of lipstip. “Our ambition is to give them the precision and speed they deserve, via completely sovereign AI.”
By automating part of the legal analysis tasks, the startup therefore seeks to modify the very economics of the sector. The objective is not only to reduce file processing time, but to allow firms to manage more volumes without proportionally increasing their human resources.
Supported by Hélioparc, the Nouvelle-Aquitaine Region and Aquiti Gestion, lipstip also carries out its research work in partnership with the University of Pau and Pays de l’Adour as well as with CY Tech.
