Will Big Data Keep Us Honest?

Data can tell if you're lying, but is this always a good thing?


A few weeks ago I ordered a vintage cycling jersey which never arrived, the sender had sent it to the correct address, it had been signed for but not by anybody I knew. After opening a case with the delivery company and agreeing compensation after two weeks of conversations, I got an email from the sender saying that the package had arrived back with him. I had a moral dilemma, as it could be resent and I would end up with the compensation and the jersey, but Big Data kept me honest. The same day I received an email from the delivery company saying that their data showed that the parcel had arrived back with the sender.

Although in the end I didn’t have a decision to make, it is a prime example of how data is being used to keep us honest.

With our actions being tracked in almost everything we do, it gives us less chance to be dishonest. There are even examples of insurance companies tracking how forms are being filled out to detect dishonesty. This is an interesting concept as if they are filling out a form to try and get the cheapest possible price, this will be shown. They may say that the car is kept in a garage rather than on the street to reduce the premiums and test this through filling out the forms with multiple variables to bring down the premiums.

Big Data could also be used to track lies through social media, something that is having an increasing influence on the way that companies are perceived or even how markets react. A prime example of this is when the press association had their Twitter account hacked and a tweet was sent claiming that a bomb had gone off in the White House and President Obama had been injured. With this rumour, markets dropped by 140 points before recovering when it came to light that it was a hoax.

There are applications that can track this kind of lie and make sure that they are flagged efficiently in order to make sure that this kind of thing does not happen in future. It is important for markets and companies that this kind of lie can be identified and nullified in order to make sure that the damage is as small as possible. Companies like Pheme are going to be important in this as they can use algorithms and real time analytics to help prove or disprove any potential rumour.

It is not only at company and national level that this could have an impact though, we have seen that data and sensors could help us to identify if somebody is lying to you on an individual level. Through voice intonation, body language and pupil dilation, it is possible to use data to identify if the person in front of you is being dishonest. It is something being used at the moment in an experiment for border security by the University of Arizona, where automated interviews can scan through previous visa applications and forms, whilst also tracking body language to identify whether or not they are being truthful.

A similar system is also being used to track whether registered sex offenders are sticking to their probationary conditions, through tracking their body language under questioning.

We may even see similar systems being used on our phones or wearable technologies in the future. For instance if we had a phone conversation with somebody, their truthfulness may well be tracked through the intonation of their voice or if you are talking to them through a programme which shows their face, could track their facial movements to see if their are telling the truth.

There is certainly a potential for this for the greater good, but when it filters down to a personal level it could become fairly invasive. Although I do not tend to lie to people, the fact that you could always be tracked by whoever you are speaking to to find out if you are lying is a scary thought. If you are trying to tell a white lie to protect somebody’s feelings then it can be picked up. One thing is for certain though, that it could make for some very awkward conversations in the future. 


Read next:

Working At The Boundaries Of Aesthetics And Inference