Facial Recognition: The Good, the Bad and the Unfortunate

Early 2020 has seen big changes to the implementation of facial recognition software around the world. These changes are not without criticism and a rising tide of concerned citizens and watch groups worry the technology could pave the way to mass surveillance as seen in other nations, including China.

There are compelling arguments on both sides of the facial recognition debate. Here, we take a look at three benefits and three concerns around facial recognition technology.

The benefits of facial recognition technology

Safety and security

By far the most persuasive argument for the adoption of facial recognition is enhanced safety and security. Proponents point to cases where the tech has been put to use finding missing children and elderly individuals. One case from New York highlights how quickly facial recognition can help solve crimes: police apprehended a man less than 24 hours after he threatened a woman with rape at knifepoint. And in a widely shared story, a man has been reunited with his biological family after being abducted as a toddler


Image: Uunsplash

Airports are turning to facial recognition increasingly. In the US, the Department of Homeland Security predicts that the tech will scan 97 per cent of travellers by 2023. The idea is twofold: people are less likely to commit a crime if they are aware that being watched and tightened security at airports will lead to fewer serious terrorist incidents.

The convenience factor

When we once thought of facial recognition as an easy way to tag ourselves in Facebook photos, the convenience factor could increase exponentially if we rely on biometric scanning in other areas of our lives. Offering a quick and intelligent verification process, it’s possible that facial recognition can be used for payments and other processes in the future.

Seamless integration and low costs

Although facial recognition is sophisticated technology, it is remarkably easy to implement the tools into existing systems. Companies and law enforcement agencies needn’t go to a lot of effort to start using the technology and most facial recognition applications are compatible with existing security systems. As an added plus, the technology is not expensive to run or purchase. As a case in point, the Washington Post reports that controversial app Clearview AI costs less than US$8 per month.

The drawbacks of facial recognition technology


Image: Pixabay

An invasion of privacy

Across the world, an increasing number of people and organisations are joining the fight against facial recognition technology. Their primary concern? The blatant infringement on basic civil liberties. Already San Francisco has become the first city in the world to ban the technology amidst privacy concerns and activists elsewhere are asking their home cities to do the same.

While being scanned at an airport is one thing, having one’s biometric data taken without consent while walking to the shops or the local park is quite another. Unlike online privacy where you could to some extent control by using tools like VPNs, it seems impossible for people to avoid the widespread use of facial recognition.

Use without check

Despite assurances from the government that facial recognition technology would only be used in specific law enforcement contexts, these assurances have already been broken. Here we reach the crux of the issue when it comes to facial recognition as a tool: without proper checks and accountability in place, agencies and/or businesses are free to put the technology to use as they see fit, further infringing on individual privacy.

An example of this comes from New Zealand where police illegally used Clearview AI’s troublesome facial recognition program without seeking clearance from the Privacy Commissioner.

In a data-driven world where user data is now more valuable than oil, there is also a very real concern that the technology will be used and abused by hackers, whether they should have access to it or not. As a Gizmodo reporter discovered, downloading Clearview’s app was not particularly difficult.

Misidentifications

Facial recognition technology works on either a one-to-one basis (where one image is compared against one other) or on a one-to-many basis (where one image is compared against a databank of images to find a match).

An issue is that facial recognition technology is not foolproof and has demonstrated an unfortunate racial and gender bias: it seems to have issues differentiating between people of colour and women. Misidentification could lead to individuals jailed while still “in-trial”, a process of considerable judicial length for many.

Amongst the ongoing implementation of facial recognition technology, it is little wonder people are turning to ways to ensure their privacy. Facial recognition is here to stay but the state must find a way to balance the pros against the cons and implement its use in a way that guarantees rights and liberties.

No comments:

Post a Comment