OnGrid Urges Financial Sector to Reinforce Video KYC Against Deepfake Fraud

As India accelerates its digital onboarding of financial customers, it also faces an invisible threat - synthetic identity fraud powered by deepfakes. A 700% rise in such attacks in 2023 alone has prompted security experts and fintech leaders to urge for immediate reinforcement of India’s Video KYC protocols. OnGrid, a leading digital trust platform, is sounding the alarm and offering a tech-forward solution that not only meets compliance but actively prevents identity spoofing.

“The future of financial fraud is synthetic. Without native deepfake detection and anti-spoofing layers, companies are only protecting themselves on paper—not in practice,” said Vaibhav Yadav, Product Head at OnGrid.

Deepfake technology has become commoditized, with over 2,000 facial manipulation tools, 1,000 voice-cloning apps, and dozens of KYC bypass kits available on underground markets. These tools allow imposters to masquerade as legitimate users, create synthetic identities, or launch replay attacks often with little technical expertise required. What once took expertise and resources can now be executed in minutes at negligible cost.

Despite RBI’s digital push, gaps remain: no regulatory definition of deepfake fraud, no mandatory reporting of spoofing incidents, and underinvestment in anti-spoofing infrastructure.

OnGrid’s Smart Video KYC directly addresses this challenge by embedding:

●       Active and passive liveness detection

●       Real-time facial and voice-video synchronization checks

●       AI anomaly detection in biometrics

●       Device and geolocation audit trails

“Anti-spoofing must be built into protocols, not bolted on as a patch,” added Vivek Kumar, Product Manager at OnGrid.

OnGrid is urging regulators to classify synthetic identity fraud, mandate anti-spoofing in Video KYC, and enable sector-wide threat intelligence sharing. Initiatives like OnGrid’s advanced deepfake & spoof-resistant video KYC demonstrate that robust defenses are possible, if institutions are willing to invest ahead of the threat curve.

The Rise of a Commoditized Threat

Deepfake technology, once the domain of hackers and sophisticated threat actors, has gone mainstream. A growing underground ecosystem now offers fraudsters a ready-made toolkit:

●       2,000+ software tools for manipulating faces in real-time or recorded videos

●       Over 1,000 applications capable of cloning voices with alarming accuracy

●       Nearly 50 ‘bypass kits’ tailored to defeat conventional video KYC protocols

These tools allow imposters to masquerade as legitimate users, create synthetic identities, or launch replay attacks often with little technical expertise required.

What was once an expensive, time-consuming process now costs less than a smartphone and can be executed in minutes.

Financial Sector: Exposed and Ill-Equipped

Despite the RBI’s push for digital onboarding, experts say India’s financial ecosystem remains vulnerable to deepfake-based identity fraud.

While banks, NBFCs, and fintechs have widely adopted video KYC protocols, their defenses have not kept pace with the sophistication of the threats they face. Key vulnerabilities include:

●       No formal regulatory classification of deepfake or synthetic identity fraud

●       Absence of mandated reporting for video KYC failures linked to spoofing

●       Underinvestment in anti-spoofing infrastructure, especially among smaller institutions

●       Rapid scaling of video KYC without proportionate enhancements in security

Anatomy of a Deepfake Attack

Experts outline multiple vectors by which deepfakes infiltrate KYC journeys:

●       Face Spoofing: AI-generated facial overlays or hyper-realistic masks used to impersonate real users

●       Voice Cloning: Fraudsters emulate a target’s voice to pass audio-verification layers

●       Replay Attacks: Previously recorded video or audio is played to defeat basic liveness checks

●       Synthetic Identities: Entirely fabricated personas, combining AI-generated face, voice, and documents

In most cases, these techniques are sophisticated enough to bypass first-generation KYC systems that lack active detection mechanisms.

Why Anti-Spoofing Must Become Standard

The solution, experts argue, lies in moving beyond compliance checklists toward a multi-layered trust infrastructure that proactively detects fraud.

At the forefront of this effort is OnGrid’s Smart Video KYC, which integrates:

●       Active and passive liveness detection

●       Real-time facial movement analysis

●       Voice-video synchronization validation

●       AI-based anomaly detection in biometric patterns

●       Audit trails with time-stamped geolocation and device fingerprints

This tech-first approach does not merely meet RBI’s KYC mandates—it goes several steps further, enabling institutions to prevent onboarding fraud before it happens.

“Anti-spoofing must be embedded at the protocol level, not just layered on top as a cosmetic add-on,” said Vivek Kumar, Product Manager at OnGrid.

Policy Catch-Up: The Regulatory Gap

Even as fraudsters become more agile and well-resourced, regulation has struggled to keep pace. Currently, there is no statutory mandate for identifying or reporting deepfake-related onboarding failures.

Experts are calling for the following measures:

1. Formal classification of synthetic identity fraud within the regulatory reporting framework
2. Mandatory anti-spoofing checks in video KYC guidelines
3. Shared threat intelligence platforms to track deepfake tools and attack patterns across the sector
4. KYC audit norms that go beyond document validation and assess biometric authenticity

The underlying message from experts and technology providers is clear: compliance alone is not security. In an age of synthetic fraud, trust must be engineered into every step of the customer onboarding journey.

Initiatives like OnGrid’s advanced deepfake & spoof-resistant video KYC demonstrate that robust defenses are possible, if institutions are willing to invest ahead of the threat curve.