The term biometrics, is thrown around a lot when talking about identity in the digital age, when people say ‘biometrics’ the mind immediately starts to imagine some sci-fi esque big data program, scanning your face and reading your emotions!
There’s a lot of misconceptions of what ‘Biometrics’ are and how they are relevant in the digital age. This blog post attempts to outline what Biometrics is, and the use case for identification.
What are biometrics, and what’s a biometric check?
‘Biometrics’ is the blanket term for distinguishable human characteristics – how we walk, talk, look, act, and behave.
It follows, then, that a biometric check is simply matching two sets of characteristics to see if they’re a match. For example, when you walk through an arrivals gate, and they hold up your passport to see if one biometric (your face) matches another (the picture on the passport), or similarly when a bank clerk checks your driver’s license when you go into the branch to do your banking.
Why are biometrics important?
When it comes to information security as it pertains to Personally Identifiable Information (PII), there isn’t really anything more important than ensuring you’re dealing with the right person. Anything you can do to increase your confidence that the person you think you’re dealing with is legitimately that person, is a thing worth doing.
So what makes a good biometric? A good biometric is something that’s hard to fake – something that narrowly defines the individual as being uniquely that individual.
This isn’t new, either: Consider the iPhone. 4-digit codes to unlock a phone were easy to steal, and easy to impersonate, so back in 2013, Apple added a layer of biometric security in the form of a fingerprint reader to unlock the phone. Then in 2017 with the iPhone X they introduced facial recognition as an improved biometric. Why is facial recognition so important? Because in addition to a face being unique, it’s also captured “live” on a video stream.
There’s another important reason biometrics are important, and that is: Regulations demand them.
The 2009 AML/CFT act legislates that a reporting entity take reasonable steps to satisfy itself that the information obtained under section 15 (identity requirements) is correct.
It is entirely reasonable, then, to collect a likeness of the individual being verified as a biometric when performing a remote verification. This causes significant confusion on what good standards are for biometric identification.
Many companies claim to have ‘biometrics’ with no underlying technology or facial recognition software to back up these claims.
Facial Recognition and liveness – Facial recognition is the most natural means of biometric identification. The face recognition system does not require any contact with the person – this continues to become more and more important in a technology lead, post COVID19 world where in person contact with customers is becoming less frequent.
Good facial recognition algorithms are getting extremely accurate with Artificial Intelligence. According to a 2018 National Institute of standards and Technology (NIST) study, the industry has made massive gains in facial recognition accuracy in the last five years (2013- 2018). NIST found that 0.2% of searches, in a database of 26.6m photos, failed to match the correct image, compared with a 4% failure rate in 2014. It’s a 20x improvement over four years.
The best forms of Facial recognition technology include a ‘liveness test’ which involves a short video being taken to ensure the person is not using false imagery or tampering with the software in any way.
How does First AML help?
An in-person biometric check (such as the airport arrivals passport check) is great, but not always practical. What if the person you are verifying isn’t available to come into your office, and you aren’t able to meet them at theirs? What if they’re not in the same city as you, or even the same country? What if the whole world is in lockdown through a global pandemic?
Here’s where First AML’s remote verification process can help.
When your customer completes a First AML identity verification form on their mobile phone or computer, in addition to gathering all the data you require for AML purposes, we use robust 1000 point facial recognition technology to provide the customer with a worldclass remote verification experience. We then use matching algorithms to make sure their face matches the picture on the document. You can have confidence, then, that the person completing the form is in possession of the document, and that it is legitimately theirs.
If you want to learn more about First AML get in touch at firstaml.com/contact