Last Updated: 10 September 2026
Arlequin AI, a Paris based startup developing artificial intelligence for complex decision making, is working on a different type of neural network for governments and large organisations. Founded by Hugo Micheron and Antoine Jardin, the company focuses on systems that can analyse large amounts of connected information while allowing users to trace results back to the original evidence. Its work is aimed at situations where understanding how data points relate to each other matters as much as the final result.
Europe is also putting more public resources into artificial intelligence and the infrastructure behind it. In June 2026, the European Commission introduced a technological sovereignty package covering areas including semiconductors, cloud infrastructure, software and AI. One month later, it announced plans for up to seven AI Gigafactories, supported by up to €10 billion in EU and national funding and expected to attract at least €20 billion in private investment. European Commission technology sovereignty initiative EU AI Gigafactories announcement
Arlequin has now raised €28 million in Series A funding. Redalpine and OTB Ventures co led the round, with Bpifrance’s Defence Innovation Fund also participating. Existing investors Vsquared Ventures and 10x Founders increased their stakes, while Xavier Niel and ZEBOX, the startup accelerator backed by CMA CGM, joined as new investors.
Arlequin is developing topological neural networks, known as TNNs. These AI models are designed to learn from the way information is connected, including relationships involving several elements at the same time. The focus is on connections inside complex data.
One example comes from counterterrorism investigations. Arlequin says its platform can analyse millions of data points from multiple seized devices and identify links between people, locations, communications and events. TNN based models are intended to extend this process as datasets become larger and relationships become harder to identify.
The company says the same approach can be used in security and defence, criminal investigations, fraud, money laundering, cybersecurity, information integrity and AI system security. Arlequin is also designing the architecture to require significantly less computing power, which the company says could reduce its dependence on energy, advanced semiconductors and computing infrastructure while keeping results traceable to their source.
“AI sovereignty is not only about where models are built or data is hosted. It is about having control over the technologies that increasingly underpin our most critical decisions. Europe needs AI systems it can own, understand and trust. Arlequin is building a fundamentally different paradigm of AI, designed to uncover relationships in complex data without predefined ontologies, bias or hallucination, while keeping results traceable and verifiable. This is critical not only for governments, but also for enterprises where reliability is non-negotiable. We are proud to co-lead this round and excited to back Hugo, Antoine and the Arlequin team as they enter their next phase of growth.”
Jeremy Teboul, OTB Ventures
The new capital will support the development and training of Arlequin’s proprietary TNN models, along with international expansion and commercial deployment. The company plans to grow the team working on the technology and continue research collaborations with INRIA, CNRS and the Max Planck Institute, as well as the universities of Oxford, Cornell, Princeton and UC Santa Barbara.
Expansion is another part of the plan. Arlequin already has offices in London and Berlin alongside its Paris headquarters, and it plans to open an AI laboratory in Silicon Valley by the end of 2026. The company is also increasing its commercial activity outside France.
Arlequin says its technology is already deployed with governments and large organisations across Western and Eastern Europe. It is currently active in four European countries.
Micheron and Jardin founded Arlequin in Paris in 2024. Micheron, the company’s CEO, holds a PhD and specialises in jihadism and geopolitical instability. Jardin, its CTO, previously worked as a CNRS research engineer in data science and human behaviour.
The company’s existing platform can work with several types of information, including documents, transactions, video and operational data. It analyses relationships between those sources so users can identify patterns and connections across large datasets. The system is designed for cases where users need to understand both the result and the evidence behind it.
Arlequin employs around 50 people, including 35 engineers and 15 PhDs, researchers and data scientists. Its technology is used in areas including security and defence, criminal and terrorist network investigations, fraud, money laundering, judicial investigations, cybersecurity and information integrity.
“We believe the next advances in AI will come from fundamentally different architectures, not simply from making models larger and feeding them more data and energy. LLMs have transformed language and content generation, while world models have extraordinary potential for physical AI. With Arlequin’s topological neural networks, we are opening a new field of artificial intelligence for problems that existing architectures are not designed to solve efficiently.”
Antoine Jardin, CTO and co founder of Arlequin AI
The €28 million round is backed entirely by European investors. Redalpine and OTB Ventures co led the financing, while Bpifrance’s Defence Innovation Fund participated alongside them. Vsquared Ventures and 10x Founders increased their existing stakes, and Xavier Niel and ZEBOX joined the investor group.
The Series A will finance further work on its TNN models while supporting the company’s team growth and wider commercial deployment.
“Most of the capital in AI right now is chasing the same architecture and hoping scale solves everything. Arlequin is pursuing a genuinely different architecture: developing topological neural networks as a distinct architectural approach, rather than relying solely on larger models and more compute. What sets the company apart is that this research is the direct engine for a product that is already being used by governments and enterprises, a combination of fundamental science and commercial traction that remains rare in Europe.”
Maria Portoles, Investment Partner at Vsquared
