Let me ask you something that doesn’t get discussed enough in the boardrooms.
Your institution has invested heavily in credit models. You’ve refined risk scoring, tightened underwriting parameters, and built decisioning engines that many would call best-in-class. But when was the last time you examined what happens in the thirty seconds before any of that kicks in?
That’s the gap. And it costs you more than you think.
What We Got Wrong About Digital KYC
When the Reserve Bank of India introduced Video KYC in 2020, most institutions responded by doing what institutions typically do they treated it as a compliance obligation to be operationally absorbed, not a strategic capability to be leveraged.
That was a mistake. And the consequences are now showing in terms of fraud numbers, write-off ratios, and portfolio quality at institutions across the country.
Here’s the uncomfortable truth: We digitized documentation. We didn’t digitize our identity. Aadhaar OTP and PAN validation tell you a person exists. They tell you nothing about who they actually are, whether the profile in front of you is synthetic, or whether the intent behind the application is legitimate. We built faster pipes, but we didn’t fix what was flowing through them.
Between 2020 and 2025, digital lending volumes crossed $350 billion in India. In the same period, identity fraud and synthetic profiles grew two to three times over almost entirely exploiting the gap between documentation and genuine identity verification. The attack surface didn’t grow despite digitization. It grew because of it.
The KYC Architecture Most of Us Are Still Running
Walk through a typical onboarding stack at a mid-to-large NBFC or bank today. You’ll find documents validated, OTPs confirmed, databases queried. Tick, tick, tick. And then, several layers downstream; risk models finally begin their work analyzing financials, bureau data, repayment histories.
By that point, a sophisticated fraudster has long since passed through the gate.
The structural flaw isn’t in risk models. They’re doing exactly what they were designed to do. The flaw is that they’re positioned too late in the sequence. Risk evaluation begins after identity has already been accepted and in a world of synthetic profiles, that’s an irreversible error.
No underwriting model, however sophisticated, can recover from an identity that should never have been onboarded in the first place.
What Video KYC Actually Enables When Done Right
The institutions that are pulling ahead aren’t using Video KYC as a compliance checkbox. They’re using it as the first layer of their risk architecture.
Think about what a well-designed Video KYC interaction captures: liveness confirmation, facial consistency, behavioral signals, geolocation context, response patterns under structured questioning. That’s not just identity verification that’s a behavioral dataset. And it’s the highest-signal, least-contaminated dataset you’ll collect across the entire customer lifecycle.
The shift, when you see it clearly, is this: KYC stops being about confirming identity and starts being about interpreting it. That’s not a semantic distinction. It has direct downstream implications for fraud rates, portfolio quality, and the efficiency of every risk model that follows.
Institutions that have made this shift are seeing fraud intercepted before account creation not detected post-disbursement. That’s a fundamentally different cost structure. Collection pressure drops. Write-offs compress. And every Video KYC interaction feeds a learning system that gets sharper with volume.
The Strategic Implications for CXOs
This is where I want to be direct with you, because I’ve seen too many well-resourced institutions miss this transition by treating it as an IT implementation rather than a business architecture decision.
First: onboarding is now a core risk function. If your KYC process sits inside operations and reports through a compliance lens, you have an organizational problem, not just a technology one. The CRO and CDO need direct ownership of onboarding architecture and the data that flows from it.
Second: AI cannot be bolted on. Video KYC without embedded intelligence face matching, anomaly detection, behavioral scoring is expensive, inconsistent, and ultimately no more effective than what it replaced. The intelligence must be native to the process, not layered on top of it.
Third: the traditional sequence of KYC, then underwriting, then the decision is collapsing. In the next phase of digital lending, that sequence becomes a single real-time decisioning layer. Your competitive position will be determined by how far upstream your risk of intelligence reaches and how fast it operates when it gets there.
Finally: the institutions expanding aggressively into Tier 2 and Tier 3 markets are doing so without proportional physical infrastructure. The constraint has shifted from branch density to digital orchestration. Video KYC, done well, is what makes that economics work.
Where This Is Going
The next major divide in Indian financial services will not be between institutions with better credit models. It will be between institutions that understand identity as a risk function and those that still treat it as a compliance formality.
As platforms like Think360 push further into alternate data intelligence and real-time risk monitoring, Video KYC becomes the entry point into a continuously learning system one that improves with every interaction, refines its fraud signals with every application, and feeds better intelligence into every downstream decision.
The winners in this market won’t be the ones who approve applications fastest. They’ll be the ones who decide, with the highest certainty, who deserves to be in the funnel at all.
That decision the one that determines everything that follows is now being made in the first thirty seconds of a video call.
The question is whether your institution is using those thirty seconds or wasting them.
