Check new clients – with evidence for the supervisor.

Why AML client checks take so much manual work — and how traceable AI evidences every step.

1 The problem: the signals sit in many data silos.

The German Anti-Money Laundering Act requires general due diligence for every new business relationship: identify the contracting party and establish who is behind it economically — down to the natural person.1 If you identify a higher risk, enhanced due diligence applies, for example where high-risk third countries are involved.2 The information you need sits in many places: company name and seat in the client file and the commercial register, the beneficial owners in the Transparency Register,3 names on the consolidated EU financial sanctions list.4 Each source looks harmless on its own. Only together do they show whether you need to look more closely. And later, supervisors or auditors ask: why did you onboard this client?

2 Why checklists and generic AI don't solve this reliably.

An Excel checklist records that something was checked — but not what the sources add up to. Querying registers and lists by hand means copying results from one window to the next, and the connection gets lost. And an AI chatbot? The US standards institute NIST states that fabricated statements are inherent to the design of such models.5 Without access to your client file and the registers, a generic model cannot check the specific client anyway. Even when the AI searches real documents, the error rate only drops — it does not disappear.6 For a decision you have to justify to your supervisor, that is not enough.

3 The approach: client, registers, sanctions lists and law in one graph.

c:node links your client file with the c:node Graph — commercial register, Transparency Register, sanctions lists and the Anti-Money Laundering Act as a network of points and edges: a knowledge graph. Research calls this combination of pattern recognition and firmly checked rules neuro-symbolic AI; it delivers answers you can justify and trace.7 Across the graph, c:node reasons several steps ahead — client → holding → director → sanctions list entry — which plain text search cannot do.8 Every step stays tied to its source.

4 In practice: scattered sources become a well-founded decision.

The same question for which you query registers and lists one by one today, c:node answers in a single pass — and discloses every statement:

5 Auditable for supervisors and auditors — and sovereign.

The Anti-Money Laundering Act requires you to record the information gathered during the client check and the result of the risk assessment, and generally to keep it for five years.9 c:node stores every finding with its source and legal basis. This creates an audit trail you can show your supervisor, such as BaFin, or your auditors — without disclosing the inner workings of the model. Your team still makes the decision; c:node prepares it. From 10 July 2027, the EU Anti-Money Laundering Regulation (EU) 2024/1624 will also largely apply directly in all member states;10 you connect new obligations in the graph instead of rebuilding checklists. And because c:node runs in the EU cloud or on your own servers, client data stays in the EU or in your own data centre.