Are you welcoming the right individuals, or inadvertently granting access to the wrong ones?
In pursuit of a "Zero Touch" experience, the BFSI sector has invested heavily in removing friction. While one-click accounts and instant approvals were the goal, by 2026, this approach had made digital interfaces more vulnerable. The industry now faces a paradox: the frictionless customer journey has become the main entry point for Video Injection Attacks and Deepfake Fraud.
The 2026 Fraud Landscape: Beyond Traditional Masks
The threat landscape has evolved. Legacy systems were designed to prevent "Presentation Attacks" — such as using photos or static masks. Today, fraudsters can bypass these measures without appearing on camera.
- Video Injection Attacks: Attackers employ virtual cameras and emulators to inject high-fidelity, AI-generated video directly into the system stream. The system perceives this as a legitimate live feed.
- The Compliance Gap: The RBI's 2025-26 updated KYC Master Directions have raised standards. Face matching alone is insufficient — regulators now require liveness verification that resists synthetic manipulation.
- The Latency Trap: Many banks rely on "Active Liveness" checks (asking users to blink, smile, or turn their heads). These are easily mimicked by AI and often frustrate genuine customers, resulting in a 40% drop-off rate in V-CIP sessions.
Introducing "Smart Friction": A Strategic Solution
Consider friction not as a UX failure, but as a dynamic security tool. Smart Friction remains invisible to 95% of genuine users while presenting significant challenges to the 5% who trigger risk signals.
1. Passive Liveness (The Invisible Shield)
Instead of requiring users to perform actions, the system conducts Passive Liveness checks. It analyzes skin texture, micro-movements, and background light reflections discreetly. This process is faster for users and significantly more difficult for deepfakes to replicate.
2. Device Integrity & Environment Cues
Before the video call begins, a smart CRM checks for indicators such as emulator usage, active virtual camera drivers, or suspicious IP velocity. These signals enable the system to adjust friction levels in real time.
3. From Data Entry to Risk Intelligence
In this model, Video KYC agents are empowered to focus on analyzing behavioral cues and human-level anomalies that AI may overlook, rather than manually entering data that OCR can process.
The SimpleWorks Edge
- Edge-AI Processing: We address latency by processing initial liveness heuristics on the user's device, preventing session timeouts and ensuring a smooth experience.
- Smart Orchestration Engine: Automatically trigger Smart Friction steps only when high-risk indicators are present, keeping conversion rates high and fraud rates near zero.
- Audit-Ready Governance: Every session, every rejection, and every AI decision is logged in a tamper-proof audit trail that meets the latest RBI norms for V-CIP compliance.
