All Posts
Guides6 minute read

How bunq Protects Your Account from Fraud

bunq
All Posts
Guides6 minute read

How bunq Protects Your Account from Fraud

bunq
All Posts
Guides6 minute read

How bunq Protects Your Account from Fraud

bunq

Table of contents

bunq processes billions of transactions. Behind each one, a set of systems works to keep your money where it belongs: catching unusual patterns, verifying payment details, and stepping in when something doesn't add up.

Most of the time, none of this is visible. Payments go through, everything works as expected. When something looks off, the system catches it before you've had a chance to notice.

Here's what's running while you get on with your day.


How fraud detection works

Every transaction you make runs through a fraud detection model that analyzes more than 500 data points in real time. Not just the amount: the location, the time of day, the merchant, and how the payment fits your usual behavior.

The model uses hundreds of decision trees competing against each other simultaneously. When a transaction fits the profile of known fraud, it's flagged within seconds. When it doesn't fit any known profile but still looks off, a second layer catches it.

That second layer is an unsupervised deep learning model—meaning it doesn't work from a list of known fraud types. It runs across every payment the supervised model clears, looking for complex patterns that don't match any labeled fraud type. When something new surfaces, it gets labeled and fed back into the supervised system. The two models strengthen each other continuously.

This is how bunq keeps up with fraud schemes that didn't exist last year. The system doesn't rely on a fixed rulebook. It updates as new patterns emerge.


Why some payments get paused

When a payment raises a safety concern, Safety Shield makes sure your money doesn't move until you've had a chance to check it. The payment is paused for 24 hours. During that window, you can review it and cancel if something seems off. 

Fraudsters rely on urgency. They push you to confirm before you've had time to think. The hold exists to remove that pressure, and it can't be bypassed by you or by bunq. That's intentional. If the payment was legitimate, it clears when the window closes. If it wasn't, you cancel it in the meantime.


How bunq detects remote-access attempts on your device

Remote-access fraud works by convincing someone to install screen-sharing software, like AnyDesk, so a scammer can see and control what's on their phone. It's effective because it operates from inside the device, where most security layers don't reach.

bunq detects when remote-access or screen-sharing software is active. If there's reason to suspect it's being used to access your account, bunq limits the account immediately and lets you know. The detection runs at the app level, so you don't have to know what to look for.


Why biometric login means only you can approve payments

Every time you open the app or confirm a transaction, biometric login ties the authorization to your body, not just your device.

bunq uses your device's own biometric hardware for this. Face ID maps more than 30,000 data points before it authenticates; modern fingerprint scanners process around 10,000. Neither stores a photo of your face or fingerprint. What's held is an encrypted mathematical representation of the pattern, which can't be reversed into the original image.

Your biometric data is processed and stored entirely within your device's secure hardware. bunq never receives it.

If biometric authentication isn't available, your PIN works as a fallback. But day-to-day, biometric login means that even with physical access to your device, approving a payment still requires you.


How IBAN Name Check catches misdirected payments in the moment

Social engineering often works at the payment step. A fraudster provides an IBAN that looks valid but belongs to a different account holder. The money leaves before anyone checks whether the name matched.

bunq's IBAN Name Check runs a name-to-IBAN verification on every SEPA payment before you confirm it. If they don't match, you see a warning. You can proceed or cancel, but the mismatch is flagged at the moment you can still do something about it, not after the transfer has gone through.

This runs automatically on every outgoing transfer. You don't need to enable it or remember to check. The verification is built into the payment flow.

For a closer look at individual safety features you can manage yourself, Is bunq safe and trustworthy? goes deeper on account controls in the app.


Getting help when something goes wrong

Finn is bunq's AI support assistant, and your first point of contact when something needs attention. Whether a payment's been flagged or you just have a question, Finn resolves 84% of cases directly, day or night, without a wait.

Finn has context on your account, so the first response is relevant to your situation rather than a generic prompt to describe your problem. Most cases get resolved without needing to go any further.

Over the past six months, Finn has taken on more than 40% of the questions that previously needed a human guide. When you do need one, for a disputed payment, a fraud recovery, a complaint that needs an actual decision, you get there faster.


When your case goes to an expert, automatically

Some situations need a person: a fraud recovery decision, a formal complaint closure, anything where the outcome requires someone to own it.

bunq routes those cases automatically. You don't request an escalation or restart the conversation with someone new. The system identifies when a guide needs to take over and gets you there, with your context already carried across.

At most banks, the experience goes like this: you explain the problem, get a reference number, and wait weeks for a response. At bunq, the conversation is the resolution. You reach out, explain what's happening, and by the time you close the chat your case is typically sorted. 9 out of 10 times, in under 5 hours from start to finish.


Final thoughts

The security layer at bunq is made up of several systems, each covering a different part of the picture: detection before a payment processes, a hold when something looks wrong, support that gets you to the right answer without friction, device-level authentication, and a name check at the exact moment a transfer can still be stopped.

None of it asks anything different of you. The systems run quietly and surface when something needs attention.

bunq also uses AI to help keep you safe beyond these features. Learn more on the AI & Safety overview.

Share this post

Table of contents

bunq processes billions of transactions. Behind each one, a set of systems works to keep your money where it belongs: catching unusual patterns, verifying payment details, and stepping in when something doesn't add up.

Most of the time, none of this is visible. Payments go through, everything works as expected. When something looks off, the system catches it before you've had a chance to notice.

Here's what's running while you get on with your day.


How fraud detection works

Every transaction you make runs through a fraud detection model that analyzes more than 500 data points in real time. Not just the amount: the location, the time of day, the merchant, and how the payment fits your usual behavior.

The model uses hundreds of decision trees competing against each other simultaneously. When a transaction fits the profile of known fraud, it's flagged within seconds. When it doesn't fit any known profile but still looks off, a second layer catches it.

That second layer is an unsupervised deep learning model—meaning it doesn't work from a list of known fraud types. It runs across every payment the supervised model clears, looking for complex patterns that don't match any labeled fraud type. When something new surfaces, it gets labeled and fed back into the supervised system. The two models strengthen each other continuously.

This is how bunq keeps up with fraud schemes that didn't exist last year. The system doesn't rely on a fixed rulebook. It updates as new patterns emerge.


Why some payments get paused

When a payment raises a safety concern, Safety Shield makes sure your money doesn't move until you've had a chance to check it. The payment is paused for 24 hours. During that window, you can review it and cancel if something seems off. 

Fraudsters rely on urgency. They push you to confirm before you've had time to think. The hold exists to remove that pressure, and it can't be bypassed by you or by bunq. That's intentional. If the payment was legitimate, it clears when the window closes. If it wasn't, you cancel it in the meantime.


How bunq detects remote-access attempts on your device

Remote-access fraud works by convincing someone to install screen-sharing software, like AnyDesk, so a scammer can see and control what's on their phone. It's effective because it operates from inside the device, where most security layers don't reach.

bunq detects when remote-access or screen-sharing software is active. If there's reason to suspect it's being used to access your account, bunq limits the account immediately and lets you know. The detection runs at the app level, so you don't have to know what to look for.


Why biometric login means only you can approve payments

Every time you open the app or confirm a transaction, biometric login ties the authorization to your body, not just your device.

bunq uses your device's own biometric hardware for this. Face ID maps more than 30,000 data points before it authenticates; modern fingerprint scanners process around 10,000. Neither stores a photo of your face or fingerprint. What's held is an encrypted mathematical representation of the pattern, which can't be reversed into the original image.

Your biometric data is processed and stored entirely within your device's secure hardware. bunq never receives it.

If biometric authentication isn't available, your PIN works as a fallback. But day-to-day, biometric login means that even with physical access to your device, approving a payment still requires you.


How IBAN Name Check catches misdirected payments in the moment

Social engineering often works at the payment step. A fraudster provides an IBAN that looks valid but belongs to a different account holder. The money leaves before anyone checks whether the name matched.

bunq's IBAN Name Check runs a name-to-IBAN verification on every SEPA payment before you confirm it. If they don't match, you see a warning. You can proceed or cancel, but the mismatch is flagged at the moment you can still do something about it, not after the transfer has gone through.

This runs automatically on every outgoing transfer. You don't need to enable it or remember to check. The verification is built into the payment flow.

For a closer look at individual safety features you can manage yourself, Is bunq safe and trustworthy? goes deeper on account controls in the app.


Getting help when something goes wrong

Finn is bunq's AI support assistant, and your first point of contact when something needs attention. Whether a payment's been flagged or you just have a question, Finn resolves 84% of cases directly, day or night, without a wait.

Finn has context on your account, so the first response is relevant to your situation rather than a generic prompt to describe your problem. Most cases get resolved without needing to go any further.

Over the past six months, Finn has taken on more than 40% of the questions that previously needed a human guide. When you do need one, for a disputed payment, a fraud recovery, a complaint that needs an actual decision, you get there faster.


When your case goes to an expert, automatically

Some situations need a person: a fraud recovery decision, a formal complaint closure, anything where the outcome requires someone to own it.

bunq routes those cases automatically. You don't request an escalation or restart the conversation with someone new. The system identifies when a guide needs to take over and gets you there, with your context already carried across.

At most banks, the experience goes like this: you explain the problem, get a reference number, and wait weeks for a response. At bunq, the conversation is the resolution. You reach out, explain what's happening, and by the time you close the chat your case is typically sorted. 9 out of 10 times, in under 5 hours from start to finish.


Final thoughts

The security layer at bunq is made up of several systems, each covering a different part of the picture: detection before a payment processes, a hold when something looks wrong, support that gets you to the right answer without friction, device-level authentication, and a name check at the exact moment a transfer can still be stopped.

None of it asks anything different of you. The systems run quietly and surface when something needs attention.

bunq also uses AI to help keep you safe beyond these features. Learn more on the AI & Safety overview.

Share this post

Table of contents

bunq processes billions of transactions. Behind each one, a set of systems works to keep your money where it belongs: catching unusual patterns, verifying payment details, and stepping in when something doesn't add up.

Most of the time, none of this is visible. Payments go through, everything works as expected. When something looks off, the system catches it before you've had a chance to notice.

Here's what's running while you get on with your day.


How fraud detection works

Every transaction you make runs through a fraud detection model that analyzes more than 500 data points in real time. Not just the amount: the location, the time of day, the merchant, and how the payment fits your usual behavior.

The model uses hundreds of decision trees competing against each other simultaneously. When a transaction fits the profile of known fraud, it's flagged within seconds. When it doesn't fit any known profile but still looks off, a second layer catches it.

That second layer is an unsupervised deep learning model—meaning it doesn't work from a list of known fraud types. It runs across every payment the supervised model clears, looking for complex patterns that don't match any labeled fraud type. When something new surfaces, it gets labeled and fed back into the supervised system. The two models strengthen each other continuously.

This is how bunq keeps up with fraud schemes that didn't exist last year. The system doesn't rely on a fixed rulebook. It updates as new patterns emerge.


Why some payments get paused

When a payment raises a safety concern, Safety Shield makes sure your money doesn't move until you've had a chance to check it. The payment is paused for 24 hours. During that window, you can review it and cancel if something seems off. 

Fraudsters rely on urgency. They push you to confirm before you've had time to think. The hold exists to remove that pressure, and it can't be bypassed by you or by bunq. That's intentional. If the payment was legitimate, it clears when the window closes. If it wasn't, you cancel it in the meantime.


How bunq detects remote-access attempts on your device

Remote-access fraud works by convincing someone to install screen-sharing software, like AnyDesk, so a scammer can see and control what's on their phone. It's effective because it operates from inside the device, where most security layers don't reach.

bunq detects when remote-access or screen-sharing software is active. If there's reason to suspect it's being used to access your account, bunq limits the account immediately and lets you know. The detection runs at the app level, so you don't have to know what to look for.


Why biometric login means only you can approve payments

Every time you open the app or confirm a transaction, biometric login ties the authorization to your body, not just your device.

bunq uses your device's own biometric hardware for this. Face ID maps more than 30,000 data points before it authenticates; modern fingerprint scanners process around 10,000. Neither stores a photo of your face or fingerprint. What's held is an encrypted mathematical representation of the pattern, which can't be reversed into the original image.

Your biometric data is processed and stored entirely within your device's secure hardware. bunq never receives it.

If biometric authentication isn't available, your PIN works as a fallback. But day-to-day, biometric login means that even with physical access to your device, approving a payment still requires you.


How IBAN Name Check catches misdirected payments in the moment

Social engineering often works at the payment step. A fraudster provides an IBAN that looks valid but belongs to a different account holder. The money leaves before anyone checks whether the name matched.

bunq's IBAN Name Check runs a name-to-IBAN verification on every SEPA payment before you confirm it. If they don't match, you see a warning. You can proceed or cancel, but the mismatch is flagged at the moment you can still do something about it, not after the transfer has gone through.

This runs automatically on every outgoing transfer. You don't need to enable it or remember to check. The verification is built into the payment flow.

For a closer look at individual safety features you can manage yourself, Is bunq safe and trustworthy? goes deeper on account controls in the app.


Getting help when something goes wrong

Finn is bunq's AI support assistant, and your first point of contact when something needs attention. Whether a payment's been flagged or you just have a question, Finn resolves 84% of cases directly, day or night, without a wait.

Finn has context on your account, so the first response is relevant to your situation rather than a generic prompt to describe your problem. Most cases get resolved without needing to go any further.

Over the past six months, Finn has taken on more than 40% of the questions that previously needed a human guide. When you do need one, for a disputed payment, a fraud recovery, a complaint that needs an actual decision, you get there faster.


When your case goes to an expert, automatically

Some situations need a person: a fraud recovery decision, a formal complaint closure, anything where the outcome requires someone to own it.

bunq routes those cases automatically. You don't request an escalation or restart the conversation with someone new. The system identifies when a guide needs to take over and gets you there, with your context already carried across.

At most banks, the experience goes like this: you explain the problem, get a reference number, and wait weeks for a response. At bunq, the conversation is the resolution. You reach out, explain what's happening, and by the time you close the chat your case is typically sorted. 9 out of 10 times, in under 5 hours from start to finish.


Final thoughts

The security layer at bunq is made up of several systems, each covering a different part of the picture: detection before a payment processes, a hold when something looks wrong, support that gets you to the right answer without friction, device-level authentication, and a name check at the exact moment a transfer can still be stopped.

None of it asks anything different of you. The systems run quietly and surface when something needs attention.

bunq also uses AI to help keep you safe beyond these features. Learn more on the AI & Safety overview.

Share this post

Related reading

Explore more tips and stories from bunq.

Fake bank calls and vishing: how to spot them and what to do
Life4 minute read

Fake bank calls and vishing: how to spot them and what to do

Someone calls claiming to be from your bank. They know your name and say your money is at risk. Here's what's actually happening, and what to do.

Moving to France from the US: Visas, Money, and Expat Banking
Life7 minute read

Moving to France from the US: Visas, Money, and Expat Banking

Moving to France from the US? Learn about visas, budgeting, healthcare, taxes, driving rules, and expat banking in France with bunq.

Safe online banking: practical tips
Finance5 minute read

Safe online banking: practical tips

Tips for keeping your bank account safe online, from setting up 2FA to spotting phishing, controlling your cards, and what bunq handles automatically.

Related reading

Explore more tips and stories from bunq.

Fake bank calls and vishing: how to spot them and what to do
Life4 minute read

Fake bank calls and vishing: how to spot them and what to do

Someone calls claiming to be from your bank. They know your name and say your money is at risk. Here's what's actually happening, and what to do.

Moving to France from the US: Visas, Money, and Expat Banking
Life7 minute read

Moving to France from the US: Visas, Money, and Expat Banking

Moving to France from the US? Learn about visas, budgeting, healthcare, taxes, driving rules, and expat banking in France with bunq.

Safe online banking: practical tips
Finance5 minute read

Safe online banking: practical tips

Tips for keeping your bank account safe online, from setting up 2FA to spotting phishing, controlling your cards, and what bunq handles automatically.

Related reading

Explore more tips and stories from bunq.

Fake bank calls and vishing: how to spot them and what to do
Life4 minute read

Fake bank calls and vishing: how to spot them and what to do

Someone calls claiming to be from your bank. They know your name and say your money is at risk. Here's what's actually happening, and what to do.

Moving to France from the US: Visas, Money, and Expat Banking
Life7 minute read

Moving to France from the US: Visas, Money, and Expat Banking

Moving to France from the US? Learn about visas, budgeting, healthcare, taxes, driving rules, and expat banking in France with bunq.

Safe online banking: practical tips
Finance5 minute read

Safe online banking: practical tips

Tips for keeping your bank account safe online, from setting up 2FA to spotting phishing, controlling your cards, and what bunq handles automatically.

Ready for easier banking?

Learn how bunq uses machine learning, Finn, Safety Shield, biometric login, and IBAN Name Check to protect your money around the clock.

Ready for easier banking?

Learn how bunq uses machine learning, Finn, Safety Shield, biometric login, and IBAN Name Check to protect your money around the clock.

Ready for easier banking?

Learn how bunq uses machine learning, Finn, Safety Shield, biometric login, and IBAN Name Check to protect your money around the clock.