What is the Future of IDs?

The situation in Syria is quite concerning. You are probably sitting at home in a non-war torn situation, thinking that is really sad for them, but how does that impact me?  

On the one hand, the refugee crisis is not an issue for any single country. As an example, a big part of the 2016 US presidential debate revolved around Syrian and other refugees. There was a lot of acrimony and even some fear. How do we know who these people are? Is it possible there’s terrorists that are using this system to get into another country? And a lot of it boiled down to this idea of identity. So these questions, very realistically, do permeate borders. 

On the other hand, this is not an isolated experience in Syria or places outside of Syria. These types of technologies, biometrics in particular, are really permeating both national and private ID systems for everyone across the globe. If you have crossed any border in recent years, you’d see the increase of biometrics being used to control who is coming in and out of the country. 

In fact, it’s not just at the border. Look at your phone, your tablet, and your computer. Do you log into them with your fingerprints or face scan every day? 

Biometrics as a component of identity will increase over time. And how we want to employ or control that will be important. We’ve already moved from just using fingerprints to facial recognition. What is coming next? The underlying technology of that is biometric by nature. And if we open that box, all kinds of questions pop out: who owns your biometric data, who’s monitoring it, how is it being stored and what they can do with it? 

Those who are considering these questions may be more reluctant to adopt these new technologies. Some people are even terrified of these new identification systems, and they want to hold onto the passport, a physical representation. And to the opposite, there are other people who are quickly shifting over, embracing digital identity which they consider safer and more accurate. 

While that might be true overall, what if biases and errors find their way into such systems? One example is Apple. When Apple first introduced their facial recognition software, there was actually bias in the initial programmes where minority groups’ faces were not identifiable. In an extreme instance, people with different faces could unlock somebody else’s phone, just because the algorithm couldn’t recognise their faces being unique.  

As we discussed in Module 4, that’s the issue with bad data that gives bad results. The failure percentage might be as low as one or two percent, but such failure might have huge remenfications. Among the refugees in Syria, one or two percent would have represented hundreds of thousands of people. And the number would be huge if it’s across the globe. If you have a bad password, you can change the password. But if your fingerprint doesn’t match up, how do you go through and actually rectify that process?  

Again we acknowledge that biometric data statistically provides fairly low failure rates, but such failure cases should not be omitted. Common practices of digital identity verifications are combinations of biometrics and other forms of authentication. Such a combination creates a bit of redundancy, just in case you encounter either technology issues or a situation where certain aspects of your biometrics have changed. So, we think the redundant approach of having multiple forms of identity to create the full identity portfolio is going to be the mainstream for quite some time.