Polish Fund OTB Ventures Invests in AI That Tracks Money-Laundering Networks Where ChatGPT Falls Short

French startup Arlequin AI has closed a €28 million Series A funding round, co-led by Polish venture capital fund OTB Ventures. The money will fund development of technology based on topological neural networks that detects the hidden connection structures behind money laundering, terrorism financing and disinformation, areas where classic large language models like ChatGPT struggle to keep up.

What Sets the Technology Apart

Arlequin AI built a platform called HuDEx that, instead of analyzing individual documents or messages sequentially the way large language models do, maps structural patterns across millions of interconnected data points. In other words, the system doesn't read the content of a single message, it looks for answers to who is contacting whom and how, in order to uncover networks invisible to traditional tools.

The topological neural networks behind HuDEx are designed to run on far less computing power than the transformer models underlying ChatGPT or Claude. The company's founders also stress that the system's outputs are explainable and auditable, which matters in court proceedings, where an algorithm's decisions need to be justified before a judge.

Today's tools are great at writing text. We're opening an entirely new chapter: solving problems where existing systems are simply helpless. - Antoine Jardin, CTO of Arlequin AI

Who's Behind the Company

The company is led by Dr. Hugo Micheron, a French political scientist specializing in jihadism, winner of the 2023 Geopolitics Prize, and an advisor to successive French presidents on terrorism issues. The technology is led by Antoine Jardin, a former CNRS research engineer and data scientist focused on social-behavior analysis, who helped co-write France's 2018 National AI Strategy and worked on the Jean Zay supercomputer.

That combination of expertise, terrorism and geopolitical knowledge on one hand, advanced mathematics and computing infrastructure on the other, explains why investors see Arlequin AI as more than just another startup building a wrapper around existing language models.

Applications and Clients

The system is already used commercially by governments and large organizations across Western and Eastern Europe, supporting criminal investigations and counterterrorism operations. The company lists money-laundering detection, financial fraud prevention, cybersecurity, court proceedings, and identifying disinformation campaigns among its areas of application.

Arlequin AI currently operates in four European countries, with offices in London and Berlin, and maintains research partnerships with institutions including France's INRIA and CNRS, Germany's Max Planck Institute, and Oxford, Cornell, Princeton and UC Santa Barbara.

The Polish Fund's Role

OTB Ventures, a Warsaw-based deep-tech-focused fund, co-led the round alongside Swiss firm Redalpine. The fund focuses on companies in the space, enterprise automation and AI, fintech infrastructure, and cybersecurity sectors, and has invested in nine new companies over the past twelve months.

For Poland's venture capital market, taking part in such a large European deep-tech round signals that local funds are increasingly confident about co-leading financing for advanced technology startups from outside the country, rather than simply adding to rounds led by Silicon Valley investors.

All the investors in the round share one thing in common: they are exclusively European, which the company and the funds describe as a deliberate choice, reflecting the continent's growing focus on technological sovereignty in security and defense.