Forget the mask, wear a niqab

Your face is now the key. With or without N-95 masks.

In the near future, your face will be the only way to access an ATM, get on a plane, bus or car, enter a building, or check into your hotel room. Possibly even your own home.

Feeling secure? Well, don’t. Worried about privacy? You no longer have any. Remain anonymous? No way. In public or in private. Especially the former.

Thanks to facial recognition software, our endless fascination with all things digital will soon become a nightmare -- a society in which anyone can spy on you. Your friend, your co-worker, your family, even a perfect stranger. Not to worry, you can do the same.

Though there have been persistent calls to enact legislation to protect citizen privacy, the emphasis so far has been on protecting simple text data, such as account numbers and passwords.

A far greater threat to citizen privacy has been ignored. Huge searchable image databases have been created by many governments and private companies from images willingly posted by social media aficionados or ‘captured’ by miniature cameras embedded in any number of innocuous-looking devices including phones, CCTVs and smart TVs, doorbells, locks and speakers.

These image databases, when used in conjunction with face recognition software, can be used to locate ‘a face in a crowd’, so to speak. These days, you don’t even have to own a computer or phone in order to be tracked -- your mere presence in a camera-rich environment is all that is required.

The pen is no longer mightier than the sword. The camera is.

The utility of facial recognition systems stems from their presumed ability to recognise a particular face in a sea of faces, be it at a cricket match, protest rally, wedding, or Zoom meeting. It has been well documented that people have been denied jobs, visas and housing based on their social media postings. The same thing is guaranteed to happen, even more so, when image databases are mined to extract information. Key players in this data mining operation are Clearview (the world’s leading facial recognition software company), Facebook, Microsoft, Google and Amazon.

For AI’s learning algorithms to work properly, an enormous amount of data needs to be collected and manually input. This means most of the newly created AI jobs will be highly repetitive, mind-numbing data-entry jobs requiring workers with marginal technical skills to analyse, classify, and label audio, video and text data according to scripted company requirements (that determine what constitutes human faces, hate speech, pornography, violent imagery, etc). Are these jobs all that different from assembly line work at garment and shoe factories? Once the AI programmes are deemed adequately trained, the entire analysis/classification/labelling process can be automated, and the workers terminated.

Friedrich A Hayek, co-winner of the 1974 Nobel Prize in Economics for his study of the interconnected nature of economic, social and institutional phenomena, remarked that, in a totalitarian State, the goals of the planner become the goals of the governed, as do the means needed to achieve said goals. That the growth of populism across the globe seems to track the growth of the IT industry is no accident.

Why would any country be interested in cracking down on dubious IT company practices when its larger interests are immensely benefited from data held by these companies? When monopolistic IT companies get in bed with both authoritarian and supposedly democratic governments, their objective is to create the perfectly obedient citizen, one ready to consume material goods and State propaganda with equal facility.

In an October 2020 NYTimes op-ed piece ‘Why We Must Ban Facial Recognition Software Now’, authors Evan Selinger and Woodrow Hartzog wrote, “Stopping this technology from being procured — and its attendant databases from being created — is necessary for protecting civil rights and privacy. But limiting government procurement won’t be enough. We must ban facial recognition in both public and private sectors, before we grow so dependent on it that we accept its inevitable harms as necessary for ‘progress’.”

Facial recognition systems, as currently deployed, have great difficulty recognising people of colour and women. I am sure that the IT giants with AI research and development activities in India are focused on solving this problem
with almost two billion coloured faces – free of cost!

Want to protect yourself from the coronavirus and the prying AI in the sky, both at the same time? And, as an added benefit, from drones with missiles?

Wear a niqab, in public and in private.