USE CASE · COMPLIANCE AND ONBOARDING
Registry data is collected via partners, and the customer confirms instead of filling in. A few open questions are enough, for companies and private individuals alike. The AI compares the answers with the registers and delivers a structured KYC file with deviations, input for the risk class and an audit trail.
Example with a fictional company.
THE KYC FILE
Choose an example. On the left, the customer sees a simple dialogue. On the right, the file that compliance receives grows, with every answer compared with the registers and every step logged with source and time.
PRIVATE INDIVIDUALS · NEW ACCOUNTS
A common scheme today: an elderly person is defrauded outside the bank. The money is then passed on through seven or eight accounts using Swish, Sweden’s mobile payment system, and from there to a foreign account or into crypto. The accounts are often opened by money mules.
Some money mules are recruited only after the account has been opened. That is why ongoing follow-up is needed too.
According to Finansinspektionen, the Swedish financial supervisory authority, an elevated risk of money laundering is not in itself a reason to refuse or close an account. The bank must make an individual assessment, try measures and help the customer provide what is missing. That takes a dialogue, and documentation that shows why.
Each customer is assessed individually. The bank must test whether the risk can be managed with measures before refusing.
The bank must explain what information is needed. The right follow-up questions do that in the same dialogue.
Refusing entire customer groups hampers the fight against crime. AMLA and the EBA are to issue guidelines on this by 10 July 2027.
Source: Finansinspektionen, “Frågor och svar om grundläggande banktjänster” (reviewed October 2026).
The ordinary customer gets their account. The money mule is spotted in time.
See the dialogue with a private individualWHAT’S INCLUDED
Company details, board, owners and annual accounts are collected automatically. The customer confirms instead of filling in.
Purpose, business and repayment in the customer’s own words, in a form, chat or call.
Everything that differs between the customer’s answers and the registers, structured and explained in plain language.
Ownership and control are analysed in parallel, so that someone who controls without owning is also visible.
Vagueness, contradictions, distant language and stories that change. Signs of straw men or money mules, or of someone else steering the answers.
Screening via partners. Hits are raised to compliance together with the rest of the file.
Low, elevated or high risk under your policy, with reasoning. Compliance sets the risk class.
Source, time and the rule applied for every data point. Ready to show when the supervisor asks.
When the registers change, or the payments don’t match what the customer said, a new check is suggested.
ONE ONBOARDING, THREE WINNERS
The customer confirms what is already in the registers and only answers what is needed. Less friction, and more people who make it all the way.
Purpose, business and repayment in the customer’s own words, already summarised when the case reaches the credit team.
A structured file and full traceability in every case. Time goes to the deviations, not to collecting information.
AMLR applies from 10 July 2027. Build the process before then.
Book a demoCONTACT
Book a demo and we’ll show the onboarding with your questions and your risk policy, and what the file looks like for compliance.
Book a demo