Step 1
Consent and camera access
On the widget’s start screen the user sees why the check is needed and agrees to data processing — you set the consent text. The browser or app requests camera access once.
Passive liveness without user actions
Passive liveness is a live-presence check mode without tasks: the user does not need to blink, turn their head or follow a dot. They look into the camera, and the system looks for signs of a live face in the frames. The check becomes a step at which honest users get lost and drop out less often.
Live-presence check without tasks
Web SDK: website and WebView Result in the API and the personal account Session recording — a scenario option
The price — from the price matrix calculator at 100 000 checks a year. The time depends on the user’s camera and network
Passive liveness: a liveness check the user hardly notices
Fewer commands at the liveness step
Passive liveness answers the same question as active — is a live person in front of the camera. The difference is who does the work: in active mode the user performs commands, in passive mode the system itself looks for signs of a live face in the frames.
- 01
No commands and no countdown
The user does not read the instruction “blink twice” and does not retry because they turned their head the wrong way. They look into the camera, and the step passes without tasks or a timer.
- 02
The same protection, the work is done by a neural network
A photo, a video on a screen and a mask differ from a live face not in movement on command but in the image itself: skin texture, frame depth, edges and artefacts. These are the signs passive methods look for; when there are few of them, the check is strengthened — more on that below.
- up to 1 min per liveness step: from a few seconds
- 0 files in the widget: liveness is captured only by the camera
- suspicious session status for manual review
What passive liveness analyses while a person looks into the camera
The user sees an oval and a hint, and meanwhile the system receives one or several frames and looks in them for signs that are hard to fake all at once. Below are the signs on which passive methods are built: a description of the approach, not a list of the parameters of a specific model.
-
Face depth and volume
A live face has volume: the nose, cheekbones and ears are at different depths. In a printout and on a screen all points lie in one plane. This is one of the simplest signs for passive methods.
-
Skin texture
Pores, small irregularities, natural colour transitions. Print gives a halftone and even patches of ink, a screen gives a pixel grid, and skin smoothed by a neural network looks too clean.
-
Glare and reflections
A phone or monitor screen reflects light differently from skin, and glossy photo paper glares in a patch. Such traces are visible in the frame even on an inexpensive camera.
-
Edges, frames and hands
The edge of a sheet, a tablet frame, fingers holding a photo to the face — are frequent traces of an attack. The check looks for boundaries that a live head cannot have.
-
Natural micro-movements
Even a motionless person breathes, blinks and slightly shifts their gaze. The system does not ask for this but checks whether the series of frames contains movement that a photo and a mask do not have.
-
Traces of generation and substitution
A face created by a neural network or overlaid on someone else’s leaves artefacts: mismatched lighting, “floating” contours, odd teeth and ears. How such attacks work is described on the page about deepfake protection.
How a user goes through the check without tasks
An example — registering for an online loan from a phone through the widget. The same path — on a website from a laptop, in the WebView of a mobile app or by a link.
Step 2
A face in the oval
The screen shows an oval and a short hint “Look into the camera”. The user keeps their face in the oval while the system collects frames. There are no “blink” or “turn your head” commands.
Step 3
The step result
The system checks the frames: texture, depth, glare, edges, traces of generation. If everything is in order, the user sees “Check passed”; if not — a hint on what to fix, and they can retry.
Step 4
The result in your system
The outcome of the liveness step reaches your server by a webhook and is visible in the session card in your personal account. If recording is enabled in the scenario, the video of the session is attached to the result. What to do next — let the user in, strengthen the check or refuse — is decided by your service.
Why more honest users pass a check without tasks
-
No need to understand an instruction
“Slowly turn your head to the left, then look at the dot” — for some people this is already an exam: small print, another language, a rush. Keeping a face in an oval is clear without words.
-
Older people and people with limitations
Older people and people with impaired vision or neck mobility find it hard to perform a sequence of movements in time. Passive mode requires neither reaction speed nor precise movements.
-
Weak phones and a poor network
Tasks with animation and a series of movements are heavier for budget smartphones and slow internet. Passive methods usually need only an ordinary front camera and a few frames.
-
Fewer retries — fewer abandoned sessions
Every failed retry of a task is one more chance that the user closes the app. The number of attempts in a session is limited, and in passive mode they are wasted less often.
A seamless liveness check does not make a fraudster more honest — it removes obstacles for those who came with a real face anyway. Most often users drop out of liveness not because the system refuses but because of the tasks themselves.
How to measure where the check funnel loses people — in the article “KYC funnel: how to raise check completion and the approval rate”. The pass rate and the number of attempts at each step are visible on the personal account dashboard.
When passive mode is not enough: strengthening the check by risk level
Passive liveness is a convenient mode for mass steps, but not the only one. The more costly an error at a step, the more layers of checking are placed on top of it. The steps of the scenario are configured in the personal account.
-
Registration and low limits: passive liveness
The user creates an account, browses the catalogue, gets a small limit. Here it matters more not to lose them at the entrance: passive checking adds no tasks and confirms live presence.
Suitable for:
For marketplaces, delivery services, online schools, the first step in fintech
-
Money and documents: passive liveness + document and selfie
Before issuing a loan, raising limits or withdrawing funds, a document photo and a selfie are added to the live-presence check, and the selfie is checked for screen capture and signs of a deepfake. This confirms that the document is presented by a live person.
Suitable for:
For microfinance organisations, BNPL, wallets, car sharing
-
High risk: liveness with tasks
For a phone number change, access recovery and large transactions the scenario uses liveness with tasks: it is harder for a pre-recorded video to guess their order. How the modes differ against deepfakes — on the page “Deepfake protection in verification”.
Suitable for:
For banks, crypto services, changes of bank details
-
Disputed sessions: video recording and manual review
Enable recording of the liveness session. Sessions with the status “suspicious” are reviewed by the security team from the video rather than from a single frame, and the user is sent for a repeat check or to an office.
Suitable for:
For everyone who reviews refusals manually
Which attacks passive liveness addresses and what you get as a result
Presentation attacks are the task of liveness, digital face substitution is supplemented by selfie checks, and video-stream substitution in the widget is limited by capture only through the camera. The result of each attempt is visible in the API and in the personal account.
- A printed photo, a picture from social networks, a photo on the screen of a second phone
- A video of a person’s face on a tablet or monitor in front of the camera
- A paper or cardboard mask, a silicone mask
- A face generated by a neural network, or someone else’s face overlaid on a real one (face swap)
- Video-stream substitution: the check receives a ready-made video instead of the camera. The widget does not allow uploading a file; the remaining control is on your side
- Uploading a ready-made photo or video instead of capturing
- Weak light, the face covered by a hand or outside the oval
- Several faces in the frame — whether to block such sessions is up to you
- An outdated browser or no camera access — continue in the system browser or on another device
- The step outcome: passed or not passed, with a reason code for each attempt
- The number of attempts and the time taken — on the dashboard; if configured — the video of the session
- More on how liveness checks are bypassed — in blog article
- “Retry” — the user sees a hint on what to fix
- “Setting” — set in the scenario
- “Selfie checks” — separate selfie checks, not liveness
- “Your side” — control that stays with your service
Where passive liveness is used: registration, limit increases, going on shift
Where thousands of people go through the check every day and every extra second at the step costs part of the audience.
-
Microfinance and instalments
- A loan application from a phone in a couple of minutes: liveness sits between the questionnaire and the document and must not slow the funnel.
- Passive mode on the application, document and selfie — before disbursement. Solutions for microfinance organisations
-
Fintech and wallets
- New users with a minimum limit and limit increases later.
- Passive checking at registration, strengthening — at withdrawal. Fintech and payments
-
Marketplaces and classifieds
- Sellers and contractors register in large numbers, many from old phones.
- Live presence when onboarding a seller without extra steps. Marketplace seller checks
-
Delivery and contractor platforms
- A courier confirms that they themselves came on shift — at the warehouse or in the street.
- Passive checking before a shift — a short step. Couriers and delivery
-
Car sharing and rental
- Driver registration and a repeat check before a trip, often in a car park and in poor light.
- Liveness at registration and before renting. Driver checks for car sharing
-
App login
- Confirm that a live person, not a bot or a recording, is logging into the account — without SMS and tasks.
- On login and confirming operations by face — on the separate page
Passive liveness and the law: when a face image becomes biometrics
Liveness checks whether a live person is in front of the camera, not who they are. Whether you also use the frame to establish identity determines which rules apply to it. Below is a retelling of the rules as of October 2026, a guide for a conversation with a lawyer, not a legal opinion.
- Biometric personal data are information on a person’s physiological and biological characteristics, on the basis of which their identity can be established and which the operator uses to establish identity (Part 1, Art. 11, 152‑FZ)
- According to Roskomnadzor’s clarifications, a photo and a video count as biometrics when they can be used to establish identity and the operator uses them for this
- If the frame is needed only for the live-presence check and the face is not compared with anything, identity is not established from it. Which 152‑FZ rules to apply in this case — general or those for biometrics — is decided by your lawyer
- As soon as the scenario compares the face from liveness with the photo in a document or with a reference, it is already biometric processing
- Biometrics are processed only with a person’s written consent, except for the cases of Part 2 of Art. 11 (Part 1, Art. 11, 152‑FZ)
- Since 1 September 2025, consent to the processing of personal data is drawn up separately from other documents that the person signs (Part 1 of Art. 9 of 152‑FZ)
- A person is not obliged to provide biometrics: keep a path without a face step (Part 3, Art. 11, 152‑FZ)
- Organisations are prohibited from processing in their own systems biometrics used for identification, except for collection for placement in the EBS (clause 1 of Part 1 of Art. 15 of 572‑FZ)
- Biometrics for authentication may be processed only for comparison with the UBS vectors and stored for no more than 10 days (para. 2–3, Part 1, Art. 15, 572‑FZ)
- The law does not apply to matching that is carried out not automatically but with the involvement of an authorised official (clause 2 of Part 2 of Art. 1 of 572‑FZ)
- A company may confirm a person's action by biometrics through the system of an accredited organisation, with their consent (Part 4 of Art. 16 of 572‑FZ)
- NeuroVision liveness does not replace identification through the UBS where the law requires it: it confirms live presence, and the decision about the client stays with you
- How 572‑FZ is applied to login and confirming operations by face — on the separate page
- The retention period of session data is set in the scenario, and personal data is destroyed on request via the API
- Recording of the liveness session video is enabled in the scenario only where you need it
- Personal data are not kept longer than the purpose of processing requires (Part 7, Art. 5, 152‑FZ)
- The retelling of the rules is simplified: 152‑FZ — as of October 2026, 572‑FZ — as amended on 28.12.2024; check the current versions in ConsultantPlus. How to protect biometrics in verification processes — in the article “How to ensure the protection of personal and biometric data in KYC processes”; how the UBS works — “The Unified Biometric System”.
Why an honest user fails liveness and what to do about it
-
Light behind or in the face
A window behind turns the face into a silhouette, a lamp above gives sharp shadows. The hint “Low light” appears at once; even light from the front is best.
-
The face is not entirely in the oval
The user holds the phone too low, turns away or covers their face with a hand. The widget shows what to fix and allows a retry.
-
An old camera and browser
An image from a dirty or damaged camera, an outdated browser without camera access. It helps to wipe the lens, open the check in the system browser or continue on another device via a QR code.
-
Attempts ran out
The number of attempts in a session is limited so that an attack cannot be brute-forced. If an honest user has used them up, create a new session or move them to another path — a check with an employee or in an office.
A liveness refusal is not proof of fraud. Most often it is the light, the camera or the pose, and the user passes on the second attempt if told what to fix.
See the share of refusals by reason and the average number of attempts on the personal account dashboard: if there are many refusals because of light, change the hints on the start screen.
How much passive liveness costs
No subscription fee: completed checks are paid for. Liveness can be connected separately or as part of a KYC session with a document.
- Connect the liveness check
Liveness check
from 13 ₽ per check
Liveness only, without a document: login, going on shift, a repeat check of an already known user. 15 ₽ at 10 000 checks a year, 14 ₽ at 50 000, 13 ₽ at 100 000.
-
Recognition and anti-fraud of a document, a selfie or liveness in one KYC session. For registration with the issue of money or access to property.
- Calculate the price
Extended
on request
Everything from the “Document + liveness” package and checks against Russian databases in one KYC session. The scope and price — for your volume.
Prices — from the calculator price matrix as of the publication date. Test access — 1 month, payment only for completed checks.
Connecting passive liveness: Web SDK, link or API
The check scenario is set in the personal account and changes without a release. The liveness check opens in the widget, in the WebView of your app or by a link.
- Camera capture only in the widget
- Result in the webhook and the API
- Scenario in the personal account
- Session recording — an option
Terms of connection
1 month test, a personal manager, support under contract, payment for checks.
Glossary: live presence, presentation attack, refusal and retry
Terms you will meet in the scenario, the API response and a conversation with the security team.
- Passive liveness
- A live-presence check without tasks: one or several frames are enough, the user only looks into the camera.
- passive liveness
- Active liveness
- A check in which the user performs commands: turns their head, looks at a dot.
- active liveness
- Seamless liveness check
- A check built into the user’s path so that it is not perceived as a separate step.
- seamless liveness
- Liveness
- Confirmation that a live person, not their image, is in front of the camera at the moment of the check.
- liveness
- Presentation attack
- An attempt to deceive the camera with an artefact: a photo, a screen, a mask. More — in the article on liveness.
- presentation attack
- Video stream substitution
- Feeding a ready-made video into the check, bypassing the camera.
- injection attack
- Liveness refusal
- The step result “not passed” with a reason code. Not proof of fraud.
- liveness fail
- Suspicious session
- A session status of suspicious: it must be reviewed manually from the images and, if enabled, the session recording.
- suspicious
- Attempt
- One pass of a step in a session. The number of attempts is limited.
- attempt
- Strengthening the check
- Adding steps on top of passive liveness for risky operations: a document, a selfie, liveness with tasks.
- step-up
Licences and certificates
The project «Development of a multimodal system using artificial intelligence "Neuromed" for managing the diagnostic processes of a clinic on the basis of the "Neurovision" computer vision platform» is supported by the Innovation Promotion Foundation within the federal project «Domestic Solutions»
The Neurovision computer vision platform is registered with Rospatent (certificate No. 2023661508) and included in the Russian Software Registry (entry No. 19972 of 13.11.2023)
The right to use the Platform is granted under a licence agreement concluded by Neurovision with its customers
We will choose a liveness mode for your funnel
Tell us where you place the liveness check, how many users go through it and at which steps protection matters more — we will suggest a mode, a strengthening scheme and calculate the price.
Can't put your request into words? Book a call and our team will contact you at a convenient time
Book a call
The company is included in the register of IT companies and the register of personal data operators of the Russian Ministry of Digital Development.
Address: Moscow, Presnenskaya Embankment 12 — Moscow City business centre, Federation Tower. Nearest metro stations: Delovoy Tsentr and Mezhdunarodnaya.
Read also
Materials on liveness, attacks on face checks and protecting biometric data — in our blog.
162 25
KYC, AML and Digital Identification
Liveness Detection: What It Is, How Liveness Checks Work, and Where Businesses Need It
445 20
Anti-Fraud, Liveness and Deepfake Protection
How Liveness Detection Is Bypassed: Presentation Attacks, Replay, and Virtual Cameras
398 27
KYC, AML and Digital Identification
The KYC Funnel: How to Increase Verification Completion and the Approval Rate Without Slowing Down the Decision
446 13
KYC, AML and Digital Identification
Deepfakes in KYC: Detecting Face Spoofing and Protecting Video Verification
518 29
KYC, AML and Digital Identification
How Video Streams Are Spoofed in KYC: The Architecture of Attacks via Emulators, Virtual Cameras, and Video Stream Injection
384 16
KYC, AML and Digital Identification
How to ensure the protection of personal and biometric data in KYC processes
Frequently asked questions about passive liveness
Passive liveness is a check that a live person is in front of the camera without user actions: there is no need to blink, turn the head or follow a dot. The user looks into the camera, and the system looks for signs of a live face in the frames.
In active mode the user performs commands, and the system checks that a live person performed them. In passive mode signs of a live face are sought in the frames without tasks. Passive is simpler for the user; active is more often placed at high-risk steps. Which modes are available in your scenario, clarify when connecting.
Liveness reduces the risk of presentation attacks — photos, screen recordings, masks — but does not guarantee protection against every forgery. For large transactions, a document, a selfie with a deepfake-sign check or liveness with tasks are added on top of the passive check, and disputed sessions are reviewed manually.
With liveness the step takes from a few seconds to a minute: the time depends on the camera, the network and how quickly the user gets ready.
A front camera of a smartphone, tablet or laptop with access from a browser or WebView is needed. Poor light and a dirty lens sometimes require a retry. The camera and browser requirements — on the page on embedding the widget.
Tell them what to fix: light from the front, the whole face in the oval, no hand at the face. If the attempts have run out — create a new session or move them to a path with an employee. A liveness refusal is a reason to investigate, not proof of fraud.
Liveness checks live presence, not identity. Under 152‑FZ a face image becomes biometrics when it is used to establish identity, for example compared with the photo in a document. How to classify your scheme and which consent to take is decided by your lawyer.
Yes. The “Liveness check” package — liveness only, from 13 ₽ per check. It is placed on going on shift, login and repeat checks. For onboarding with a document — the “Document + liveness” package from 29 ₽ per session.
Through the Web SDK on a website or in the WebView of a mobile app, a KYC link without development, or the REST API: the server creates a session and receives the outcome by a webhook. The outcome is visible in the personal account, and the widget captures liveness.
Yes: a document, a selfie with a deepfake-sign check, liveness with tasks and recording of the session video are added to the scenario. For example, passive — at registration, with tasks — when changing a phone number or withdrawing a large sum. The steps are set in the personal account.
