Do you remember this scene from Minority Report?
In a futuristic depiction, many claimed that it predicted online targeted advertisements.
The kind you see all over the Internet that are based on your identity and your record
of all online activities. However, this scene depicted the future of
advertising much more precisely than that. Here you see the main protagonist walking
through a shopping mall while the displayed ads are calling him out by his name.
That movie was set in 2054 but the technology is already coming to American retail stores.
And it’s even more advanced than the movie depicted.
This is facial recognition advertising. Several US retailers, including Walgreens
and Kroger, are piloting facial recognition advertising in physical stores. A system that
uses so called “smart shelves” can already detect your age, gender or mood when you walk
by. This technology can also track your iris movements to see where you are looking specifically.
The goal is to use all available information from the cameras to target customers with
relevant ads in a similar way online advertising targets Internet users.
When you hear facial recognition, you shouldn’t just think identification – that’s the
difficult part. You should more importantly picture a mood detection system.
Your face is a very good indicator of your emotions which allow advertisers to better
tailor their ads. Your facial expressions are often times subconscious and expose your
internal reactions to certain stimuli. This mood detection is a new concept so think about
it this way. Apps on your phone can use your high definition
selfie camera to track your emotional expressions of your face as you listen to music, and suggest
you playlists that correspond to your mood. A social media platform can track your face
as you are typing that comment or text with your friends, and suggest you mood-specific
ads. When an app like this throws an ad at your
eyeballs, your facial expression can tell the advertiser a lot more about the success
of their ad than traditional engagement statistics. Emotional analytics is a way Realeyes, a startup
based in London, uses webcams to measure subconscious responses to video content. Their high-profile
clients already included Coca-Cola, Ford, Heineken, IKEA, McDonald’s, Samsung or Disney.
Its algorithm is based off six universal cross-cultural emotional states – happiness, surprise,
sadness, disgust, fear and confusion. Curiously, even the European Union supported the company
with a €3.6 million grant to better improve its product.
This technology is also going to give advertisers an omniscient power to see whether you are
actually viewing their ads, or just merely playing them in the background while doing
something else. Next generation commercials could thus require you to keep your eyes locked
in as they play before continuing to your favorite content.
The ultimate point of singularity for facial recognition advertising, is the ability to
link your face with your social media presence. Facebook, which also owns Instagram and WhatsApp,
is going to play a major role in this area. Let’s say you are standing in front of a
smart display and you see an ad. If the AI behind it thinks your facial expressions are
mainly positive about the ad and can link your face with your Facebook account, you
could be delivered a discount voucher straight to your Facebook inbox.
Facebook has one of the largest databases of pictures of people’s faces and the social
network has been implementing facial recognition since 2011. By the way, launching the Facebook’s
“Tag Suggestion” feature happened totally without user consent.
Facedeals, an Atlanta-based company, is testing a program that helps stores, bars and restaurants
identify people’s faces through their Facebook accounts. When Facedeals learns a user’s
face, its cameras will recognize them in an instant and send them customized coupons based
on their social media activity. All major technology companies are heavily
investing into the development of facial recognition, including Google, Apple, Amazon, Samsung or
Microsoft. With their massive reach on the market, it will soon become extremely difficult
for you to avoid being identified every step of your way if you want to be a functioning
individual. If you are worried about it, you are not alone.
75% of people say they would not shop at a store with facial-recognition advertising.
The same statistic also shows that 55% of consumers wouldn’t mind having their faces
scanned for ads in shops, as long as they get discounts.
With or without your consent, facial recognition will soon infiltrate every aspect of your
life. It will become a part of the new social contract. According to Market Research Future,
the global market for facial recognition will reach about $8 billion by 2022.
The technology has been silently making its way into the mainstream in almost all sectors
of modern society in both public and private spheres.
Japan began identifying shoppers’ gender, ethnicity and age as early as 2010. Today
they moved on to identifying riders in taxis and targeting them with relevant ads. Are
free rides in exchange for your face and targeted ads the logical next step in this evolution?
The entertainment industry might also experience a radical transformation during the revolution
of face scanners. Since 2013, Virgin Mobile has been working on creative campaigns where
users could choose from a variety of video scenes by blinking their eyes. Such an implementation
of interactive story telling could soon become its genre.
Because the ‘one man’s tool is another man’s weapon’ rule has no exceptions,
facial recognition has its own dystopian uses. The Chinese government is operating its massive
social score ranking system, where it uses cameras to identify its own citizens and track
their behavior. This information is then used to assign each individual a trust score, which
can significantly impact their traveling or job seeking options.
Such a social score ranking initiative would probably cause an uproar in some of the Western
countries. But others are already building such capabilities.
Countries like the UK already use security cameras on most urban public places, allowing
the government to scrutinize movements of its citizens almost everywhere. These systems
are so advanced they can identify faces even with disguises or after a surgery. In some
British schools, such cameras are already being used to take attendance.
Many airport terminals rely on facial recognition for security. E-passports utilize this technology
to speed up the queues, and it also helps officers detect fake passports with more ease.
U.S. Customs and Border Protection is in the process of installing face scanners at all
national airports. The US law enforcement is also deploying facial
recognition to identify individuals in public by cross-referencing the camera footage with
a biometric database with information on one third of Americans. The FBI’s technology,
however, scores only about 85% accuracy in identifying potential matches. This score
pales in comparison with Facebook’s DeepFace that scores 97.25% accuracy and Google’s
FaceNet with 99.63% accuracy. However, it’s not the accuracy that causes
a major concern. It’s the bias of the algorithm behind a facial recognition software.
Amazon’s Rekognition tool came under fire after it wrongly identified 28 members of
the US Congress as police suspects. But what caused even more concerns were multiple discoveries
that Amazon’s Rekognition is misidentifying women and people of color up to one third
of the time. Amazon continues to sell this technology commercially and to the police.
The problem is that any algorithm or artificial intelligence is going to be as biased as the
data set it’s using or the programmers that are coding it. One of the more commonly used
databases for this is ¾ male and 80% white. It’s important to note that facial recognition
is a system that enables classification of people by race, gender or ethnicity. Biases
against these groups implanted into the AI can result in serious discrimination.
Facebook was found guilty for letting advertisers discriminate their audiences based on their
gender, race, ethnicity or language. This could have been achieved by either showing
ads to only specifically chosen groups or by excluding certain demographics.
Facial recognition is going to allow for the same for-profit profiling but much more accurately.
Advertisers will have the capability to offer customized vouchers to certain groups while
excluding others whatever the motive. Identifying individuals before they even enter
a building will open the door for real life “deplatforming” based on what someone
said on social media, and there would be no way for the censored to avoid identification.
Currently there is no comprehensive legislative plan to govern the use of facial recognition.
When the law makers fall behind, it’s the technology that writes the rules of the game
for itself. Bahio, a coffee brand based in London, once collected faces of all passing
pedestrians during their artificially intelligent poster campaign. 42,000 people had their faces
scanned without their knowledge or consent. In the US, only Illinois and Texas have laws
requiring an explicit consent from customers when collecting their faces. But on a federal
level, there are no legal protections. And crafting them will be a nearly impossible
task. Requiring privacy policies to be present at the entry of every store using facial recognition
just puts another burden on the backs of consumers. However it will turn out to be, facial recognition
in stores will remove your ability to purchase products anonymously, even if you choose to
pay cash and refuse any loyalty membership cards. Stores and brands with posters on public
places will be able to track your social media accounts just by scanning your face.
Facial recognition is not an innovation for the common man. It is being developed by the
elites for the elites to be used on regular people. It is for this reason the city of
San Francisco has decided to ban the use facial recognition by the police and other agencies.
Other proposals around the US would ban the use of facial recognition without an explicit
consent from customers. But they have a little chance of passing.
Wherever you are, you should pay a close attention to the debate and especially to how your representatives
plan to regulate this technology. Facial recognition is incredibly advantageous to the ones who
use it, but can also be disastrously exploitative to those it’s being used upon.