The Privacy of Me

I have been thinking a lot about privacy.

There's a new store in the northwest part of our country where you can use your palm print to pay for things purchased within the store. It's not terribly complicated; you tie your online account and payment method to your palm print and when you're ready to check out you scan your items, slap your hand on a reader, and out the door you go.

Who owns that palm print?

That palm print is me. It's part of my own personal biometric signature. It is unique to me, just like a fingerprint, a retinal scan, my own face and even my gait. It can't be faked, and if you find it someplace you know that is a place I have been.

What happens if on the same day I shopped at this store, a known terrorist or international criminal or just someone of interest to the government also shopped at the store? If the government wants to know who was at the store on that day, will the management of the company who owns the store provide my palm print to the government? Can they do that without my permission? Once that palm print is a part of a record somewhere- a record I might know nothing about- it will be there forever and guarantees I can be identified simply by using it to match to a palm print collected anywhere I happen to rest my hand.

We provide data about ourselves every day, in multiple different ways. We fill out credit applications, and we buy things online. We log on to our favorite shopping website trying to find something for ourselves and in so doing provide reams of information about our shopping habits. The information gleaned from our behavior is then used to target us with advertising designed to appeal to us and increase the chance we will buy more. We can see this happen every day, as I would imagine I am not the only person to wonder how it is that things I have shopped for on Amazon or Google suddenly appear in my Facebook feed.

Data has value. The type of data might affect its value, 
but virtually all data has value.        

If you have a video surveillance system in your business, or a consumer video surveillance system in your home, it has data of value. The system itself exists with a deterrence model in mind, and this is of course a part of the value it brings to the table. If the video surveillance system happens to record a crime, then the data contained within the recording has value as well. The information contained within that video may allow you to determine who committed the crime, what they might have taken, when they entered, and can provide evidentiary value to be used in the prosecution and conviction of the criminal. The fact that the data collected is only valuable after the fact does not change the value of the data.

if these video surveillance cameras could somehow collect data that could be analyzed, perhaps in partnership with a powerful artificial intelligence, and the results of analysis used to predict behavior... Well, the value of the data provided by those video surveillance cameras would skyrocket in value.?

We tend to think of video surveillance cameras as just cameras. This sells them far short of their actual capabilities; they are more than just cameras, they are devices that collect data about us. Traditionally this data is security focused and only valuable after the fact. However, as the capabilities of artificial intelligence increase the data these cameras collect eclipse the security considerations and become valuable because they can reveal things about us as consumers.

These video cameras collect video biometrics.        

All of that junk that you often see displayed right by the registers of virtually every store on the planet comprise something called the “impulse buy aisle”. Somewhere along the line the people involved in the retail side of the world realized that while you and I were standing in line was a great time to try to get us to pick up the “oh why not” purchase of bubble gum, the latest gossip magazine, perhaps a charger for our iPhone or even a USB drive.

This wasn't done with a sophisticated artificial intelligence or analysis of customer buying patterns. Someone figured out that since we were standing in line and had nothing else to look at, it was a good opportunity to hit us with the last line of items for potential purchase.

Imagine having the ability to create an impulse buy everywhere in the store. When you walk in the items that you see appeal to you based on the time of day, your own personal demographic, the time of year, the type of shopper typically visiting the store during that timeframe, etc. We can do this now thanks to the combination of video demographics and artificial intelligence. I simply have to let my cameras collect enough data across enough of a span of time to tell me information that would allow me to create customized shopping experiences. If I know that my average shopper on a given Friday night was a young couple with no kids between the ages of 28 and 35, I could adjust displays and endcaps to appeal to this demographic as they move about the store. if that average shopper changed 2 an older demographic on Saturday morning, I could adjust the same displays to appeal to that group. Video biometrics allows me to track the physical journey my customer takes through my store, or hospital, building or school, which in turn allows me to understand more about them. Video biometrics gives me the potential ability to see where they look, and how long they look, and even further personalize their buying journey. I can do this and more with video biometrics.

Unfortunately, this same combination of video camera and 
artificial intelligence lends itself to a less ethical use.        

The video biometric systems in use within some authoritarian countries allows those governments to tell whether or not a person is a member of an ethnic minority. artificial intelligence increases the accuracy and skill of the system and provides the authoritarian regime a frightening amount of control over how they choose to persecute the aforementioned ethnic minority. Sadly, this occurs in China where that government has well over 1,000,000 Uyghur Chinese (this is an ethnic minority that happens to practice the Muslim faith) registered into a facial recognition system that tracks their movements.

Even well-intentioned video biometrics can be used to questionable ends. Within the United States there are several companies who use facial recognition and fake social media accounts to create a “friend of a friend” network with a goal of connecting with a suspect or person of interest. Our domestic news agencies have shared many stories about doorbell companies and security camera companies who share footage with law enforcement without asking permission of the system owners. Who decides whether or not your face can be used in an effort to catch a criminal? Who asks your permission to potentially involve you as an unwitting participant in a criminal investigation?

One thing seems to be clear; video biometrics is not going anywhere and artificial intelligence will continue to be used by all manner of different technologies. It seems unlikely that we could ever put the genie back in the bottle. What we can do is recognize that the use of these technologies has a good and bad, identify some of the bad use cases and work to keep them to an absolute minimum. We can also look to provide leadership and voice to an industry that is growing faster than law or regulation can possibly keep pace. We can draw on the lessons that are available to us from our recent history with social media and, rather than let the use of video biometrics and artificial intelligence grow unchecked and become bigger than is controllable as have some social media companies, work to be sure that we keep them under control.

This is why I believe Deeping Source stands alone in the 
video biometrics and artificial intelligence landscape.        

Deeping Source and our flagship product, THE SOURCE, are different. The company works within video biometrics by leveraging pre-existing video surveillance systems to provide valuable and actionable data. The company's artificial intelligence has been designed to maximize the information we can provide to a user.

The company also stands for the ethical use of biometric data. The Company's development of true data anonymization proves this, as it allows data to be pulled from video biometrics without revealing the personal identifying information of anyone on screen. The development of data anonymization as opposed to using the industry's preferred methodology, video masking, was undertaken to provide a specific way to give data without jeopardizing the individual or compromising their privacy.

So now the focus of Perspectives changes. In parallel with the podcast “The Privacy of Me” and our blog, we will be looking ever more deeply into data privacy and the world of video biometrics. Please, continue with us as we work towards a more ethical use of data.