Liveness check, or so-called “liveness detection”: how exactly does it work?

The "liveness check" mechanism is one of the key elements of the identity verification process.

Although it uses advanced biometrics and algorithms from a technical standpoint, from the user’s perspective it is a simple and quick process that typically takes only a few dozen seconds and is conducted entirely remotely.

Key Takeaways

• The liveness check is a key element of identity verification that confirms that the person on the other end is a real, living individual—not a photo, recording, or video generated using deepfake technology.
  
  
• The process is fast, remote, and intuitive; it usually takes anywhere from a dozen to several dozen seconds and requires only internet access, a camera, and simple user actions.
  
  
• Verification is based on real-time analysis of the image and behavior (e.g., facial movements, image depth, reactions to prompts), which helps detect fraud attempts.
  
  
• There are two main approaches to liveness checks: passive (without interaction) and active (with prompts for the user); however, the passive approach—which does not require the user to make any additional movements (the system then automatically analyzes the user’s movements)—is increasingly being used.
  
  
• The absence of a liveness check significantly increases the risk of fraud, impersonation, and non-compliance with regulatory requirements (e.g., AML/KYC).
  
  
• This mechanism is widely used in processes involving varying levels of risk: from banking and cryptocurrencies to electronic signatures and user onboarding.
  
  
• When implemented in accordance with standards (e.g., encryption, ISO, GDPR), a liveness check is secure and ensures a high level of user data protection.

How does a liveness check work as part of identity verification? Step by step

Below, we’ll walk through an example identity verification process using a liveness check, from start to finish.

1. Starting the verification process

After selecting an identity verification method (e.g., video verification), the user who is to undergo verification receives a link or is redirected to the process.

Everything takes place entirely online.

2. Verification of the identification document

Even before the liveness check process begins, the user is required to take and submit photos of the front and back of their chosen identification document (e.g., ID card) for data verification purposes.

The system then verifies details such as the first and last name and the PESEL number.

3. Camera access and device setup

The next step is to grant camera access.

User:

• allows the browser or app to access the camera,
• positions the device so that their face is clearly visible,
• ensures they are in a well-lit area,
• and most importantly: ensures that no other people are in the frame (the system does not approve verifications in which third parties are visible due to the risk of potential coercion that arises in such cases).

Image quality, as well as the user’s proper positioning within the frame, directly affects the effectiveness of the verification.

4. Positioning Your Face in the Frame

A frame or oval appears on the screen; you must position your face within it.

At this point, the system automatically analyzes:

• whether the face is within the frame,
• whether it is obscured (e.g., by sunglasses or a mask),
• whether the image is stable and clear.

Only after these conditions are met can you proceed.

5. Performing the instructions (actual liveness check)

Depending on the technology used, the user may be asked to perform simple actions, such as:

• turning their head to the left or right,
• blinking,
• looking in a specific direction,
• moving their face slightly closer or farther away.

These actions allow the system to confirm that:

• the image is not a static photo,
• it is not a pre-recorded video,
• the face is not obscured by anything, such as a paper mask,
• there is a real, living person on the other side.

And this is the actual liveness verification mechanism carried out as part of an active liveness check.

However, so-called “passive liveness”—that is, the automatic recording of head movements during the verification process—is being used more and more frequently. In this case, the user does not need to perform any additional commands, and the system analyzes their movements on its own.

6. Image Analysis and Fraud Detection

Throughout the entire process, the system analyzes the image in real time.

The following are checked, among other things:

• the naturalness of facial movements,
• image depth (to ensure it is not a screen),
• light reflections,
• anomalies that may indicate a deepfake or a mask,
• the presence of third parties in the recording.

This stage takes place automatically and is invisible to the user.

7. Verification Result

Finally, the user receives information about the result of the process.

The entire process usually takes anywhere from a dozen or so seconds to a few minutes without the need to contact a consultant.

What exactly is a liveness check (liveness detection)?

A liveness check is a mechanism that verifies whether there is a live person on the other end, rather than a photo, video, or deepfake.

It is part of the identity verification process, which uses methods such as recording a short video or taking a photo of the face.

This is because simply matching a face to an ID document or taking a photo or recording does not guarantee the authenticity of a person’s identity. This is especially true today, as technology is advancing at such a rapid pace, offering numerous ways to falsify such photos or recordings.

This is precisely why a liveness check is one of the key elements of modern online identity verification.

Types of Liveness Checks

Although the entire process looks similar to the user, in practice there are several approaches to detecting “liveness.” They differ primarily in terms of user convenience for the person whose identity is being verified.

Passive liveness

Passive liveness operates entirely “in the background.”

The user:

• does not need to perform any additional actions,
• only needs to make a short recording (live during the liveness check).

The system then automatically analyzes the image and detects potential signs of fraud, such as:

• screen artifacts,
• signs of a recording being played back,
• unnatural texture,
• poor/inconsistent lighting.

All of these could indicate potential attempts at fraud. This process is available, among other places, in the identity verification services offered by Autenti.

Active liveness

Active liveness, on the other hand, requires the user to interact with the system (as in the step-by-step liveness check example we presented).

The user then follows simple instructions, such as:

• “turn your head,”
• “blink,”
• “smile,”
• “look left/right.”

This allows the system to check whether:

• the person is responding in real time,
• it is not a photo or a recording,
• no attempt is being made to use a mask, a recording from another device, or a deepfake.

The simplest way to compare what a liveness check is to the Face ID feature used to unlock a phone or authorize a payment.

However, Face ID does not operate entirely “actively.” All you have to do is look at the screen, and the system automatically analyzes the geometry of your face, depth, and attention.

Active liveness goes a step further: instead of just “observing,” the system asks for specific actions (e.g., a head movement) to better confirm that there’s a real person on the other end, rather than, say, a recording or a deepfake.

But this isn’t a completely passive approach either. We aren’t taking a photo that’s then analyzed, yet we still have to look at our phone’s screen. So, naturally, we have to “engage” with the entire process.

As we mentioned above, a passive approach is increasingly being used, which automatically analyzes the user’s movements in the recording without requiring them to perform additional commands.

Why Is the Lack of a Liveness Check a Serious KYC Error?

The absence of a liveness check is a serious error within the entire Know Your Customer (KYC) process and means that the system cannot fully confirm that there is a real, living person on the other end.

This enables:

• impersonating others,
• the use of a photo or video instead of the actual user,
• attempts at fraud using deepfakes or masks.

As a result, businesses fail to meet requirements such as those under AML (Anti-Money Laundering) regulations, and they also expose themselves to fraud and financial losses.

In what situations should you use a liveness check?

You’re probably wondering when you should use a liveness detection mechanism. There are a few scenarios that will help illustrate this.

Liveness detection and UX: Does liveness verification complicate the process for the user?

Many people fear that liveness checks, especially active ones, may prolong or complicate the verification process.

However, modern solutions are:

• fast: the entire process usually takes just a few seconds,
• intuitive: the user follows simple instructions, such as turning their head or blinking, and all instructions are displayed on the phone or laptop screen,
• balanced: they ensure a high level of security without burdening the user with excessive steps.

In fact, we are increasingly seeing passive liveness verification, which is a simple process that does not require additional commands involving head movements or specific gestures.

For the user, this means spending a small amount of time performing a simple verification from home, and for businesses, it means increased conversion rates and a lower risk of fraud.

But is it secure?

Yes, liveness check solutions are completely secure when implemented in accordance with the highest standards.

For example, Autenti uses data encryption (in transit and at rest), geo-redundancy, and regular backups, and ensures the separation of user data.

The system operates in accordance with the ISO/IEC 27001:2022 standard and undergoes independent audits, meeting the requirements of, among others, the GDPR and eIDAS.

Additionally, the principles of least privilege, security monitoring, incident management, and a “privacy by design” approach are implemented, ensuring a high level of data protection and user privacy.

Liveness Check as an Element of KYC Automation

A liveness check is not only a security measure but also an integral part of the digital KYC stack, which increasingly includes:

• biometrics: facial analysis,
• OCR: document data extraction,
• AI: verification of document authenticity and user movements.

Thanks to liveness checks, it is possible to automatically determine whether verification is valid, which reduces the need for manual processing and minimizes human error.

As a result, we see faster onboarding, fewer audit issues, and greater security for institutions and customers.

FAQ: Liveness Check

Do I need to install anything to complete the liveness check?

No. In most cases, the process takes place directly in the browser or app and does not require the installation of additional software.

How long does the liveness check take?

Usually between a dozen or so and several dozen seconds. The entire verification process rarely takes more than a few minutes.

What should I do if the verification fails?

The most common causes are poor lighting, a blurry image, or a face that’s obscured. Try improving the conditions and try again.

Does the liveness check work on every device?

It works on most modern smartphones, tablets, and computers equipped with a camera, an up-to-date browser, and Internet access.

Are my biometric data secure?

Yes. The data is processed in accordance with applicable regulations (e.g., GDPR) and used solely for the purpose of identity verification.

Summary

The liveness check is a key element of modern online identity verification, ensuring that the person on the other end is a real, living person—not someone wearing a mask, using a recording, or a deepfake.

Thanks to this technology, businesses can increase confidence in their customers’ identities, reduce fraud and scams, and meet regulatory requirements such as AML and KYC.

There are two main approaches: passive, which operates in the background based on a short video clip, and active, which requires simple user actions, such as blinking or turning the head. However, the passive approach—which does not require the user to perform any additional actions—is becoming increasingly common.

Business benefits include faster and more secure onboarding, automation of KYC decisions through biometrics, OCR, and AI, reduced human error, and increased conversion rates.

The liveness check is fast, intuitive, and secure; it typically takes between a dozen and several dozen seconds, and the process requires no additional software and works on most smartphones with internet access.

See how Autenti can support your customer identity verification process, including a complete liveness check. Schedule a free demo of the Autenti eID solution.