How and Why we Made Fake ID Detection Smart

As long as there have been real IDs, there have been fake IDs. But whereas fakes were once either poorly produced, or expensive and hard to come by, today you can buy a sophisticated fake ID in a few clicks.

When we founded Scandit over a decade ago, the idea of using a consumer smartphone to detect fake IDs would have been laughable. Modern artificial intelligence was in its infancy. The cameras on early smartphones were also simply not good enough.

Move forward to today. A few weeks ago, we launched ID Validate, a new offering in the Scandit Smart Data Capture Platform. It can determine the validity of United States drivers’ licenses and state-issued IDs in just 1 second, using any smartphone.

Here’s how, and why, we made fake ID detection smart.

Why determining the validity of an ID is harder than ever before

Modern fakes are hard to detect with the naked eye. The time available to check them is also being squeezed by consumer demand for frictionless experiences.

A cashier in a bricks-and-mortar liquor store has at least 2 minutes to determine the validity of an ID. They can also assess the buyer’s body language and mannerisms while processing their items. A delivery driver dropping off alcohol ordered online, on the other hand, has about 10 seconds.

Consumers love frictionless experiences. Unfortunately, they’re also loved by under-age drinkers and criminals using false identities to steal goods.

And regulators are wising up to these new opportunities for fraud. One of the largest quick-commerce businesses in the Boston area recently lost its alcohol delivery license, because drivers were failing to validate IDs effectively.

Ask a legal or compliance team, and they’ll want the business to reduce financial, legal and reputational risk by checking every ID thoroughly.

Ask an operations team in quick commerce, last mile delivery or car hire, and they’ll have a different view. They’ll say that you can’t afford to spend several minutes determining if every ID is real or fake using manual techniques such as F.L.A.G..

Fake ID detection in 1 second

With Scandit’s technology embedded in many retail and last mile delivery businesses, we could see first-hand that manual techniques for determining the validity of IDs just weren’t working any more. Not only that, we also knew from our experience of scanning in the real world that many gig economy workers use their own smartphones, and data connection is often patchy.

This meant that ID validation software relying on high-performing hardware or an internet connection wasn’t really working either. I’ve heard first-hand stories of delivery drivers required to use an app to validate IDs at the door having to ask customers for the password to their personal WiFi.

 

The Scandit Smart Data Capture Platform already scans tens of millions of US IDs on smart devices of all kinds every year, using our advanced barcode scanning, text recognition (OCR) and object recognition software.

We decided the next step had to be to build on this existing technology and our knowledge of real-world scanning conditions to make fake ID detection smarter.

ID Validate provides easy-to-understand results in just 1 second, works on any smartphone, and doesn’t need an internet connection. It’s simple on the surface, but under the hood a huge amount of work has gone into analyzing hundreds of characteristics of a US drivers’ license or state ID – including many invisible to the naked eye.

It can be integrated into any iOS, Android or web-based application, and determines the validity of IDs on the device itself, without storing personal identifiable information.

Why smart data capture is the best technique for determining the validity of an ID

Our vision of smart data capture is to enable businesses to capture value, not just data. Empowering businesses and frontline workers with actionable, real-time information on the validity of IDs is an important next step in this mission.

Delivery drivers can now easily validate IDs. The same goes for store associates managing curbside pickup. The process itself is standardized, built into the workflow and automatically generates an audit trail.

Companies onboarding gig economy workers remotely can reduce the risk of criminals signing up using fake identities. Fake ID detection can be built into online registration for car rental, reducing the risk of theft.

I’m excited to see how quick commerce, last-mile delivery and car rental firms will put ID Validate to use. Meanwhile, we’ll continue our work uncovering more ways to make data capture smarter, and through this shift tedious work from people to technology, upskill frontline workers and empower customers.