Police in the UK are under fire for allegedly breaking ethical norms due to their use of live facial recognition technology in the fight against crime. The "lack of robust reparation" for anyone who had been hurt by the studies was a particular source of concern in a recent evaluation undertaken by the University of Cambridge on testing of the technology conducted by forces in London and south Wales. Before facial recognition technology was widely used, it was stressed that it was important to "protect human rights and improve accountability."
The Cambridge researchers aren't alone in wanting to ban police agencies from using facial recognition software. A civil liberties group in the United Kingdom known as Big Brother Watch has been running an anti-face-recognition campaign while the government considers how to regulate AI. However, outright prohibiting the use of this technology would be a bad idea.There is, in my view, solid evidence supporting a more deliberate strategy.
What is facial recognition software?
A facial recognition system is software that can identify a person in a still photograph or a moving video clip by comparing their likeness to a large database of known faces. Face detection and measurement systems are often used with ID verification services to make sure that users are who they say they are.
In the 1960s, similar systems were first developed as a type of computer application. Facial recognition systems have been around for a while, but they've recently found widespread use on smartphones and other devices, like robots. Facial recognition systems fall under the umbrella term of "biometrics" because they rely on the analysis of a person's physiological traits to authenticate the identity of the person using the system. However, the contactless nature of facial recognition systems has led to their rapid adoption, despite their inferior accuracy compared to iris and fingerprint biometrics. Advanced human-computer interfaces, video surveillance, and automatic image indexing have all made use of facial recognition technologies.
Today, governments and businesses all over the world use facial recognition technologies. Their usefulness varies; some outdated systems were abandoned for good reason. Facial recognition technology has also been the subject of debate because of concerns that its use violates citizens' privacy, leads to a high rate of false positives, promotes discriminatory attitudes based on race and gender, and fails to adequately secure sensitive biometric information. Doubts about its safety have also been sparked by the emergence of artificial media like fake news. As a result of these allegations, numerous American communities have banned the use of facial recognition technology.
wo-we Windows Hello Face Recognition Webcam for Instant Login with Windows 10, Anti-Hacking with Privacy Switch, Business IR Webcam with Dual Microphones Used for Online Conference
ANNKE C800 4K PoE CCTV Zoom Camera w/AI Facial Recognition, H.265+ Security IP Camera, 4X Optical Zoom, 164ft Night Vision, Built-in Micro SD Card UP to 128G, IP67 Weatherproof for Outdoor/Indoor
Heayzoki Facial Recognition Camera,HD Video Call Webcam,Infrared Facial Recognition Camera, Used in Access Control Systems, Entertainment Venues,etc.
Top facial recognition software includes "Kudos".
Meta has announced that it will be shutting down Facebook's facial recognition system, erasing the face scan data of over a billion users in response to rising public concern. This is going to be a huge shift in how people use facial recognition.
Law enforcement's adoption of automated facial recognition systems has increased.
To the best of my knowledge, live facial recognition is currently only used by the police in London and South Wales in the United Kingdom. Live facial recognition uses AI-driven software to compare a user's digital face to a database of known faces and evaluate how similar the two are. Even though the Manchester Police Department gave it a try in 2018, they were told to halt by the surveillance camera commissioner since they hadn't gotten the green light.
An appeals court in 2020 ruled against the deployment of the technology by the South Wales Police. The court said that the legal framework for the force's deployment gave them almost complete freedom in how they used the technology. In this case, the fact that the police already told the public about the information didn't change how the case turned out. This is called "overt operational deployment." Despite many police agencies' claims that they are not currently using face recognition technology, this ruling does not prevent them from doing so.
Any police department in the UK may use facial recognition technology under the current legal framework.
The Metropolitan Police of London are increasingly relying on facial recognition technology to help them find missing people, suspects, witnesses, and victims. Scan after scan of faces has been made using a police car or street pole mounted facial recognition camera during public events and city plazas. Citizens who enter the zone of recognition are alerted to the deployment, unless doing so would compromise police operations or be an emergency.
Between February 2020 and July 2022, the equipment was installed by the Met in a total of eight locations, one of which being Piccadilly Circus. They reportedly examined over 150,000 facial images, which led to nine arrests but also eight blunders in which they misidentified the wrong person.
Cons and pros.
Even while there have been significant developments in facial recognition in recent years, such as the capacity to function in real time, the technology is not yet perfected in terms of accuracy. Allegedly, 228 people in New Jersey were wrongfully arrested between January 2019 and April 2021 thanks to the use of non-real time face recognition technology. One Black American was wrongfully incarcerated for 11 days due to a mix-up in identification. Using a bogus ID can lead to all sorts of problems, such missing flights and having to answer questions from the police.
There are some groups that are hit far harder than others. In 2019, a study in the United States found that the risk of misidentification was ten times lower for male faces than for female faces, and it was one hundred times higher for black and Asian faces than for white faces. This technological constraint could be used to continue the disproportionate number of stops and searches that people of color experience at the hands of police officers.
Facial recognition may or may not be a part of the answer.
Additionally, covert, permanently installed facial recognition cameras could be installed by police enforcement. For instance, this could help the government quash protests more effectively. The threat of state surveillance is one reason many people want this technology banned, and there is a lawsuit pending against Russia in the European Court of Human Rights over similar practices.
Despite this, utilizing facial recognition has various benefits. It can help authorities track down terrorists, missing children, and other dangerous criminals and potentially hazardous people.
We have adopted a capitalist system that heavily depends on big corporate spying, whether we like it or not. The United Kingdom and the United States both have a significant concentration of CCTV cameras. A London resident's personal information is recorded an average of 300 times every day, and police can use this information without a warrant. Moreover, the fact that huge corporations in the technology sector have access to virtually all of our personal data is cause for grave concern. We shouldn't be worried about live facial recognition because we already tolerate so much surveillance.
A smarter approach.
Rather than explicitly prohibiting the use of covert facial recognition or any other form of this technology, I support a legislative law to limit when it can be implemented. For one, it can be used by British law enforcement to keep tabs on those currently on their watchlists, which may include people who have just been charged with minor offenses. Furthermore, there is no universal set of criteria used to decide who gets included.
Under the new rule that is being proposed for the EU, only those accused of crimes with a possible punishment of more than three years could be exposed to facial recognition testing. That seems like a reasonable place to set a boundary.
Second, the deployment must be approved by a court or other impartial agency. When requesting this kind of permission, it's important to consider whether or not the deployment would be reasonable given the specified police mission. Now a superintendent or higher in the Met Police Department must give the go-ahead for a search warrant to be executed, and they must use their discretion to determine whether or not the search warrant's scope is proportional.
What's more, we need ethical standards that can be audited and are completely open to public scrutiny at every stage of the process, from planning to implementation to maintenance. For instance, if a photograph of a person has been misidentified, it should be immediately taken out of circulation. The Met's official position on the topic is unclear at the moment. The Metropolitan Museum of Art is trying to use technology responsibly in other areas, but this is not enough.
Finally, developers should be legally required to train AI on a diverse enough array of populations to achieve a minimum threshold, reducing the risk of discrimination. The potential for bias will be reduced as a result of this. This kind of system should allow people to enjoy the advantages of real-time facial recognition without worrying about its downsides. It is the incorrect course of action in every aspect to outright forbid something that requires sophisticated discussion between competing interests.