identt-knowledge-vision-biometric-verification-systems-online

Government Issued ID Verification

I have discussed the general idea of Identity Verification. Almost all identity verification scenarios involve the verification of some kind of government issued ID to establish the true identity and know exactly who we are dealing with in the digital space.

Now, let’s face it: there is no simple and “solve it all” government issued ID verification in the digital space. Good news is that if you do it right, it can be far more secure and complete than doing it “on premise” e.g. doing it physically. There are over 200 countries in the world and the types of IDs issued by various governments amounts to be several thousands. 10-15 years ago, we mostly had local businesses and physical contact with a customer, and this was not really an issue. Clerks of banks or insurance companies knew pretty much the local IDs (in some cases even their owners), what they look like and how to spot a fake ID. In the digital “borderless” world, it is an increasingly difficult issue. The human brain is not wired to read security micro-prints, detect small misalignments or slightly altered font types, spot tiny obscure areas where the ID was manipulated or check the calculated logical value of an MRZ [1] code. However, Machine Aided ID Verification complemented with some AI technology can significantly improve your chances of checking an ID over the internet and extract data automatically.

Government issued IDs have gone through a major improvement thanks to increasing global tourism and business travel. About 10 years ago, standards were issued by ICAO [2] for Machine Readable Travel Documents (MRTDs). This has become widespread worldwide and almost all passports and national IDs are now qualified as fully authorized travel documents and are made according to these standards. This is why you can travel across the European Union using only an ID card. If you are interested, you can access the ICAO standards here and read some more about MRTDs here. Since then, a few companies in the world have specialized in verifying these documents. Naturally, remote verification of these documents is still a tricky thing as these documents are loaded with visible and invisible security features to make forgery very hard. It is a common practice today, that local IDs verified physically are mostly just “glanced at”. If they look ok, then it’s a pass. Fraud cases resulting from fake IDs are on the rise and almost all cases are attributed to a person checking the ID which “seems to be” a valid document. One good study on the topic can be found here.

In almost all cases, some sort of an ID is used to commit these types of crimes as machine aided remote ID verification is still not a widespread practice. This neglect comes with a cost as billions of dollars are lost to fraudsters annually. A comprehensive machine aided ID verification could eventually cut these numbers significantly. While a live person would only look at the ID with its human capabilities, machine aided verification can perform as much as 40 different types of checks including face comparison and recognition during a verification process. Further to this, data extraction becomes much easier and a fully automated procedure can be implemented with a very customer friendly interface. Though it comes with a cost for every verification made, it significantly reduces the risks as a result of ID fraud both for the business and the customer.

Needless to say, a well-thought-out ID verification procedure forms the solid background for an e-contracting scenario. Legal frameworks worldwide are increasingly flexible to allow formal contracting based on a remote ID procedure.

There are only a few companies in the world who are able to perform a remote verification service for government issued IDs as it requires a long time as well as significant financial resources to build and maintain an ID database. Artificial Intelligence technologies allow these companies to refine verification protocols and to go more in-depth into the less visible security layers of IDs. This makes remote government issued ID verification secure and will save money on the long shot.