What’s wrong with Facial Recognition? (an FAQ)

Common questions and answers on facial recognition technology from OpenMedia’s community

1. Why should I worry about my face being captured and tracked? I have nothing to hide.

Freedom from surveillance is not about having ‘something to hide’. 

It is about our right to personal privacy. We all deserve to choose what, where and with whom we share sensitive information about ourselves.

Just because you don’t think you have something to hide, doesn’t mean your surveillers agree. A grim truth is that facial recognition constantly misidentifies people who have done nothing wrong. In fact, facial recognition almost always ‘catches’ many more innocent people, than actual criminals [1, 2]. Being misidentified could have a variety of personal consequences for you, from embarrassment and inconvenience, to harassment, arrest, or physical assault. 

Facial recognition does not misidentify us even-handedly; it has been repeatedly proven 10 to 100 times more likely to misidentify women and people of colour, creating a new source of mistreatment and bias for groups who already face significant marginalization [3].

While governments and companies change – data sets are forever. Even if you trust the current government in Canada or a particular company with the responsibilities implied by including you in a recognition database, when that government leaves power, or that company sells your data, the new owner of that data can use it for very different purposes against you. Without regulation that imposes more limits on its use, your face could be used for direct marketing to you; combined with other data sets to build ‘shadow’ profiles of your behaviour on services you’ve never signed up for; or included in databases of potential criminals – all without your knowledge or consent [4,5,6]. 

And don’t forget – the more of our personal information any data set contains, the more tempting a target it is for hackers who want to sell or misuse that data.

2. Catching terrorists and criminals is critical, and I support technology that helps with that.

Preventing crimes that hurt society is important. However, policies that are justified by this goal need to demonstrate that they are legitimate– that is they actually meaningfully contribute to solving that problem – and proportional – that they represent a reasonable balance of the collective benefits provided, vs risks and damages to personal rights.

Facial recognition typically fails this test. Most current uses of facial recognition technology are indiscriminate– capturing and storing data about everyone who enters their targeted area, and/or building mass databases from wide data sets such as driver’s licenses or the open web, with no established or even suspected guilt. 

Because the technology has not been specifically considered by legislation or courts in Canada, there are few legal restrictions on how facial recognition data can be gathered, stored, and shared. And there’s no agreed standard on how many mistaken identifications are tolerable in technology used for sensitive purposes that affect people’s lives.

As a result, the facial recognition technology on the market does not meaningfully balance its benefits - often anecdotal and provided by companies that developed and sell them; 

There may well be narrowly defined circumstances in which limited facial recognition can be used in a way that is both proportional and legitimate – as with some other technologies that violate privacy. But until we have conducted a thorough, impartial, and public audit of its benefits and risks, we cannot allow fear-mongering sales pitches from the self-interested companies selling these technologies to sweep away our fundamental rights. 

3. If you don’t want to be tracked by facial recognition, you should just opt out of the spaces or apps that use it.

There is no real opting out of facial recognition – or most other digital surveillance practices.

The legality of indiscriminate and inobtrusive facial recognition technology throughout our public spaces, offline and online, means we rarely have a clear choice to opt out. Neither police nor private companies in Canada have been warning the public about their plans to deploy the technology in their community, or where they’re doing so [7]. And even if we’re informed facial recognition is in use, choosing to forego essential services like receiving a driver’s license or passing through an airport is a disproportionate cost for protecting our personal rights. 

Even when we try to opt out by avoiding a service or location we know is under surveillance, the common industry practice of creating ‘shadow’ profiles for each of us as individuals, made up from data gathered about us from other peoples’ profiles and devices, means that there is really no such thing as individually opting out of data collection and surveillance.

With that said, individual choices, like adjusting your privacy settings on online platforms and using more privacy-respecting apps, can reduce your exposure to facial recognition and tracking [8, 9, 10].

4.We’re too late to stop this tech; you can’t get in the way of progress!

It is far from too late to curtail indiscriminate facial recognition. We always have choices in what uses of new technology we normalize, and which we make unthinkable. Facial recognition is not yet widespread throughout Canada, and as laws are updated to reflect its potential, we have a real chance to decide if we allow it to become mainstream, or set tight limits on its use.

In the US, a growing number of cities have banned facial recognition technology by government agencies, including San Francisco, Oakland, and Somerville, Massachusetts. Narrower state bans have been implemented by California, Oregon and New Hampshire [11].

This is our time to decide. Citizen mobilization now can determine the government legislation and social attitudes that will shape how our society adopts this new technology. 

Want to do more?

1. Facial recognition wrongly identifies public as potential criminals 96% of time, figures reveal
2. Amazon’s Face Recognition Falsely Matched 28 Members of Congress With Mugshots
3. Facial recognition systems show rampant racial bias, government study finds
4. FACE VALUE: Marketers are using facial recognition technology to track your behavior—and personalize your shopping—but privacy advocates are worried
5. Facebook Is Giving Advertisers Access to Your Shadow Contact Information
6. Clearview AI’s Database Has Amassed 3 Billion Photos. This Is How If You Want Yours Deleted, You Have To Opt Out
7. Company suspends use of mall directory cameras running facial recognition software
8. The Complete Guide to Facebook Privacy
9. 7 Google privacy settings you should revisit right now
10. How to protect your privacy in a surveillance state
11. Facial Recognition Laws Are (Literally) All Over the Map