Skip to content
Node16.ai
ProductsRegulatoryPapersAbout usRequest a briefing →
Node16.ai / FAQ · Draft

FAQ · Draft

Frequently asked questions

The questions we hear from banks, investigators and security teams, with short answers. Ask us the rest in a briefing.

Does any bank data leave the bank?

No. Customer identities, transaction records and case evidence stay in the bank-controlled environment. What travels is protected model updates and the shared model. Case-specific sharing, when it comes with Collaborator, follows a separate, authorised disclosure path.

Can one bank use this before others join?

Yes. Sonar scores activity in the bank's own data and Investigator supports the bank's own case work. Shared learning adds a further source of intelligence as other institutions join.

Your results are per transaction. What does that mean in cases or reports?

The figures on this site count transaction alerts, before they are grouped into cases. Several alerts from the same pattern usually become one case, so your team works far fewer cases than alerts. How they group depends on the bank's case process, which we calibrate with each bank.

How realistic is the synthetic data?

Our development corpus holds about four billion transactions, four million customers and 187 money-laundering typologies, with a dense transaction network between the five simulated banks. Real Nordic interbank networks differ in density and in the mix of patterns, which is why real-bank performance is what we test next with partner banks rather than something we claim from the synthetic results.

Does the model replace the investigator?

No. Sonar ranks activity for review and Investigator prepares evidence and draft narratives. Review, escalation and reporting decisions stay with the analyst and the MLRO, and every decision is recorded in the audit trail.

How does this fit our existing monitoring system?

Sonar works alongside an existing transaction-monitoring system rather than replacing it on day one. Integration follows the bank's systems and agreed operating model, from a focused evaluation on the bank's own data to a broader deployment.

Where does training run and who operates it?

Model training and evaluation run on sovereign European supercomputing infrastructure, with product data kept in bank-scoped environments. Node16.ai coordinates the service; each bank retains responsibility for its participation and authorised use. Deployment and service arrangements are part of bank diligence.

What is the regulatory status?

Node16.ai is entering the Danish FSA's AMLR Article 75 regulatory sandbox. We are aiming for commercial clarity in Q1 2027 and developing Collaborator for risk-based information sharing under Article 75, targeting July 2027.

What about information security?

Our ISO/IEC 27001 programme is in progress and we can discuss controls, suppliers and evidence with your security, privacy and procurement teams. Company tools and the product are kept separate: bank data never enters company mail, CRM or internal AI tools.

Node16.ai

Detection. Investigation. Collaboration.
Built around the bank.

Product briefingFAQCareersData protectionPrivacyCookies
© 2026 Node16 ApS · Porcelænshaven 26, 2000 Frederiksberg, Denmark
CVR 46600207Company information