MARBLE raises €6.5 million to make financial compliance programmable

Paris-based startup MARBLE has raised a €6.5 million Series A led by SMARTFIN, with ADNEXUS, PASSION CAPITAL and 42CAPITAL participating. The company develops an open-source platform for fraud detection and AML/CFT compliance, aiming to replace the lengthy configuration cycles of legacy software.

Within a financial institution, risk rarely evolves at the same pace as the system designed to detect it. A new fraud pattern emerges, a payment channel becomes mainstream, or a regulatory obligation changes; a rule then has to be amended, a threshold adjusted, a new data source connected, sometimes an entire investigation workflow reconfigured. On paper, it is a setting. In the reality of banks, fintechs and payment institutions, it can quickly become an IT project lasting several months.

This is where MARBLE comes in. Founded by ARNAUD SCHWARTZ and PASCAL DELANGE, the startup has raised €6.5 million in a Series A round. The round was led by SMARTFIN, with participation from ADNEXUS, the investment vehicle focused on digital trust and regulated professions. Existing investors PASSION CAPITAL and 42CAPITAL also increased their stakes. The company has now raised €9 million in total.

The issue is not simply one of adding artificial intelligence to a field already saturated with technology promises. MARBLE aims to build compliance infrastructure that is closer to the way product and engineering teams operate: configurable, traceable, connected to internal data and able to evolve without restarting a transformation programme every time risk changes.

Compliance can no longer operate on an annual release cycle

Anti-money laundering and counter-terrorist financing, or AML/CFT, requires firms to know their customers, monitor transactions, screen sanctions and politically exposed persons lists, detect unusual behaviour, investigate alerts and, where necessary, file reports.

The problem is that these processes have long been supported by large software suites designed around a control logic rather than continuous adaptation. They often provide broad functional coverage, but also depend on complex configuration, lengthy integrations and substantial reliance on IT teams, systems integrators or vendors.

Financial flows have changed in nature: transactions are now instant, cross-border and multi-channel. Payment institutions, neobanks, crypto-asset platforms and fintechs have increased the pace, while fraud schemes do not wait for a project steering committee before evolving. Embedded finance infrastructures must also incorporate KYC, fraud prevention and regulatory oversight at scale.

According to MARBLE, a significant share of regulated institutions’ resources is now absorbed by compliance and manual alert handling. In many organisations, analysts still spend a substantial proportion of their time dismissing false positives, locating scattered information and rebuilding case files.

A platform designed to bring compliance, data and product closer together

MARBLE offers a real-time decision engine to monitor transactions, screen AML risks, manage investigations and change detection rules. Its distinctive feature is its open-source model and a design initially built with developers in mind.

The ambition is to enable institutions to retain control over their compliance framework: connecting their own data, integrating KYC tools or sanctions-list providers, writing or adapting detection scenarios, then deploying them with a level of traceability compatible with regulatory requirements.

This positioning also addresses an increasingly central concern in finance: where does data reside and who controls the models that use it? MARBLE offers deployment within the client’s own infrastructure. A bank can therefore, according to the company, keep transaction and investigation data in its own environment and use its own AI models or locally hosted models.

The market sits between legacy suites and new specialists

MARBLE enters a crowded market. Large banks are often equipped with solutions such as NICE ACTIMIZE, ORACLE Financial Crime and Compliance Management, SAS, FICO, BAE SYSTEMS NetReveal or FISERV. These vendors have extensive client references, teams and product portfolios; the trade-off is that they can have become too heavy for organisations seeking to change risk scenarios more quickly.

Alongside them, a new generation of specialists has emerged, including COMPLYADVANTAGE, NAPIER AI, HAWK AI, THETARAY and LUCINITY. They sell cloud infrastructure, automation, risk data and advanced analytics capabilities. HAWK AI in particular is a serious European comparable in transaction monitoring and explainable AI. In France, TOPOGRAPH illustrates another approach to compliance, centred on the quality and freshness of corporate data.

The boundary with fraud prevention has itself become less clear. FEEDZAI, FEATURES SPACE, now within VISA’s orbit, SARDINE and SEON are especially present in payments and real-time decisioning. As soon as a client seeks to bring together fraud, transaction monitoring and investigations, the categories overlap. This development echoes the evolution of online fraud detection, where platforms must combine an increasing number of signals to make real-time decisions.

MARBLE therefore does not only need to win against a software vendor. Its toughest competitor is often the existing assembly: a legacy platform, bespoke connectors, Excel files that have become indispensable, a compliance team familiar with every exception, and a systems integrator holding the system’s institutional memory. In other words, an operational debt no one really wants to risk dismantling.

The 2027 European deadline creates an opening

The regulatory timetable reinforces the appeal of this approach. The European AMLR, adopted as part of the new anti-money-laundering package, will apply from 10 July 2027. It establishes a directly applicable rulebook across the European Union, while AMLA, the new European authority dedicated to the issue, is set to strengthen sector supervision.

This deadline will not automatically create unlimited budgets, but it will compel affected businesses to assess the strength of their controls, policies, data and systems. For vendors capable of integrating without forcing a complete overhaul of the information system, the window is real. Compliance became especially sensitive following the questions raised by the Wirecard scandal over the supervision of payment players in Europe.

“All regulated businesses are now required to do more in compliance and fraud prevention with the same resources. Without automation, the equation cannot be solved,” says SAUMITRA DUBEY, partner at SMARTFIN. For the fund, MARBLE’s value lies precisely in making this automation a standard operating model directly within the client’s technology environment.

ADNEXUS, which is joining the cap table, stresses the importance of faster adaptation. “The ability to rapidly detect unusual transactions or payment orders is a major issue for financial institutions,” says DIDIER ROSSIGNOL, the fund’s chairman. The opportunity could extend beyond finance to other regulated professions, but banking, payments and fintech are where the need is currently most immediate.

AI does not remove the need to explain

MARBLE plans to use the funding across four areas: automating rule creation, improving alert handling and investigations, integrating agents into workflows, and strengthening clients’ deployment capabilities.

However, the subject must be handled with particular rigour. In compliance, AI is not assessed solely by its ability to save time: an alert must be explainable, a case file reconstructable, and a decision justifiable to an auditor, a regulator or a risk-management team. A model that reduces processing time but makes its reasoning opaque solves only part of the problem. This demand for traceability also appears in AI deployments in regulated sectors.

MARBLE says it has more than 100 institutions in production across over 15 countries, with close to 70% of its customers located outside France. The company also states that 70% of its clients selected its platform to replace an existing solution.

Its target of surpassing €5 million in ARR by 2027 means converting that adoption into enterprise contracts, with the support, security, connectors and governance expected by regulated institutions. Open source can ease initial adoption, but it cannot substitute for the trust required of infrastructure placed at the core of financial control.

The real question is whether financial institutions will finally agree to treat their control rules as a living product, rather than software they no longer dare touch once it has been installed.