Video KYC

What is Video KYC?

Video KYC is a remote customer-onboarding method in which a trained agent verifies a customer's identity through a live, recorded video call. The customer presents government-issued identity documents, performs a live face-match and liveness check against those documents, and confirms key identity details — all on camera, in real time, with the agent following a regulator-approved script. The session is recorded end-to-end, encrypted, and stored as evidence of identity verification.

Video KYC sits between fully automated electronic eKYC and traditional in-person KYC. It preserves the human assurance of in-branch verification while delivering the convenience of digital onboarding — making it a regulator-accepted alternative in jurisdictions where fully automated KYC is not yet permitted, or for higher-risk customer segments where assisted verification is preferred.

Why Video KYC matters

Three forces drove the rise of Video KYC. First, regulator acceptance — major regulators including the Reserve Bank of India, MAS in Singapore, BaFin in Germany, the FCA in the UK, and others have explicitly recognised Video KYC as an acceptable identity-verification method when conducted to defined standards. Second, customer expectation — modern customers expect to onboard from a phone or laptop in minutes, not days, and Video KYC removes the branch visit without removing the human element. Third, fraud pressure — as deepfakes and synthetic identities become more sophisticated, an assisted check by a trained agent provides a fraud-detection layer that purely automated flows cannot easily match.

How Video KYC works

A Video KYC session follows a structured workflow. The customer initiates onboarding through the institution's app or web portal, completes a brief pre-call form, and is connected to a trained agent in a queued or scheduled video session. The agent verifies the customer's location and identity, asks the customer to present a government-issued ID on camera (often capturing both sides), runs a live OCR and document-authentication check, and asks the customer to present themselves for a live face-match and liveness verification. The agent typically asks scripted questions to confirm declared information (date of birth, address, occupation), captures any required signatures or consents on screen, and records the entire session as audit evidence. The completed file — recording, document images, OCR data, biometric outputs, and agent decision — is encrypted and stored according to the institution's recordkeeping policy.

Video KYC vs eKYC

Both eKYC and Video KYC are digital methods, but they differ in human involvement, throughput, and assurance level. The table below summarises where each fits best.

Aspect	eKYC	Video KYC
Human involvement	Fully automated; no agent	Live trained agent on a recorded call
Speed per onboarding	Seconds (self-serve)	Minutes (queued or scheduled)
Cost per session at scale	Low	Higher (agent time + recording infra)
Assurance level	Strong for low–medium risk	Higher; preferred for higher-risk segments
Evidence captured	Document images, selfie, biometric scores	Full encrypted video, document images, biometric scores, agent decision
Best fit	Default low-risk segments, instant onboarding	High-risk customers, regulator-mandated assisted verification, fallback when automated checks fail

Most modern programmes deploy both — eKYC as the default for low-risk segments and Video KYC as either a default for higher-risk customers or a fallback when automated checks are inconclusive.

Video KYC for banks and fintechs

Banks were the first regulated category to adopt Video KYC at scale, primarily for current account openings, savings accounts, term deposits, and credit-card issuance where in-person verification was previously mandatory. For region-specific implementations, see our guides to Video KYC under RBI guidelines in India and the end-to-end KYC process. Fintechs and NBFCs followed quickly, using Video KYC to expand into customer segments that automated flows struggle to onboard — older customers, customers in regions with poor document quality, customers with non-standard identity profiles, or product categories where regulators require assisted verification. Insurers use Video KYC for policy issuance and surrender; securities firms use it for demat-account opening; payment institutions use it for higher-tier wallet upgrades. The common pattern is the same: Video KYC unlocks fully digital onboarding for products and segments that purely automated KYC cannot yet serve.

Liveness detection and fraud prevention

A core component of any modern Video KYC implementation is liveness detection — confirming that the person on camera is a real, present human rather than a photo, video replay, or AI-generated deepfake. Mature platforms combine passive liveness (analysing micro-expressions, depth, and texture without user effort) with active liveness (asking the user to perform random actions such as a head turn, blink, or repeat a generated phrase). High-end implementations also run real-time deepfake-detection models against the video stream and behavioural signals against the underlying device. Strong liveness detection is non-negotiable: deepfakes have moved from research curiosity to active attack vector, and a Video KYC platform without robust liveness is a liability rather than a control. For a primer on the underlying technique, see our explainer on what is a liveness check and facial liveness detection.

Regulatory requirements

Most jurisdictions that permit Video KYC impose specific operating standards. Common requirements include: trained, certified agents on the institution's payroll or under direct contractual control; sessions conducted only from secure, monitored agent locations; end-to-end encryption of the video stream; mandatory recording of the full session; geo-tagging the customer's location; live OCR and document authentication during the call; live liveness and face-match verification; capture of timestamps and randomly generated session identifiers; and retention of the complete session recording and evidence package for the period required by the local AML law (typically five years or more). Regulator-published Video KYC guidelines generally also impose throughput limits per agent and minimum agent-to-customer ratios to ensure quality.

Choosing a Video KYC platform

Key criteria when selecting a Video KYC vendor include the breadth of supported document types and country coverage; the quality of integrated liveness detection and deepfake defence (see our explainer on how deepfake can bypass biometric verification); the platform's recording, encryption, and storage architecture; agent-side workflow tools (case routing, supervisor escalation, scripted prompts); end-to-end audit trail and evidence packaging; integration into the institution's broader Video KYC API stack; and compliance with the institution's local Video KYC regulatory requirements. A well-implemented platform supports both default Video KYC flows and risk-based escalation from automated eKYC, allowing a single customer journey to step up to assisted verification when the data demands it. Many institutions pair Video KYC with a unified one-touch KYC stack and underlying biometric verification APIs, anchored against liveness detection baselines.