The Indian railways recently announced its plans of using facial recognition technology in its railway stations by the end of 2020, for better surveillance and tracking purposes. As per reports, the system is being tested at Bengaluru’s KSR city railway station, where security cameras linked to the recognition system, are being installed at all entry and exit points.
Last year, Hyderabad’s Rajiv Gandhi International Airport had launched a facial recognition system on a trial basis. Over 180 passengers enrolled voluntarily for the Centre’s DigiYatra programme, which uses a passenger’s face in the place of documents such as a boarding pass. The system has already made its debut at Bengaluru’s Kempagowda International Airport.
While the Centre’s aim is to ensure seamless, faster and safer travel for passengers, the move of introducing facial recognition systems in the country’s railway stations and airports is raising concerns of data protection and privacy.
Facial recognition is not restricted to travel alone; Telangana recently tested the technology for the first time to verify voters during the municipal elections held last month. The application, which was uploaded on mobile phones and tested in 10 polling booths, was used by the State Election Commission to counter cases of impersonation.
In another highly contentious move, at Prime Minister Narendra Modi's rally in December 2019, the Delhi police reportedly used facial recognition software to screen the crowd and identify habitual protestors and 'rowdy elements'. Every individual who walked into the rally through the metal detectors had their face screened against a list to weed out offenders.
Countries around the world have also been using facial recognition for various reasons - as databases to identify criminals, by banks to identify customers and by companies for their databases. China, for example, is the leading user of facial recognition software with around 170 million closed-circuit cameras; plans are on to increase this by another 400 million in the next three years. China has also reportedly been using facial recognition in its fight against the coronavirus, where the cameras can be used to spot people with low-grade fevers or track those who have been near a patient.
The European Union, which was the only region to oppose facial recognition and had even been considering a five-year ban on public deployment of the technology, recently backed away from calling a blanket ban. A new draft paper on artificial intelligence, as reported by the Financial Times, no longer includes the suggestion that the EU Commission should consider a five-year moratorium so that states can study their impact before rolling it out. The new draft, instead, puts the onus on members state to determine how and when they wish to allow the use of facial recognition technology.
Privacy vs security
Security cameras are ubiquitous – on the streets, building complexes, airports, shops, schools, public transport, companies, etc, so why do we fear facial recognition systems so much? Well, for one previous studies have shown that it is highly invasive and susceptible to being misused by unscrupulous elements, or by Governments for surveillance. In an experiment conducted by the New York Times using Amazon’s commercially available facial recognition service, Rekognition, the media organisation was able to identify pedestrians walking through a Midtown Park in New York using cameras installed in the park and publicly available photographs of the residents.
A study conducted in the United States in December has also confirmed what was already feared – that facial recognition systems have high chances of failure to provide accurate results. This is especially so in the case of non-whites – the study showed that the chances of showing ‘false positives’ were as much as 100 times higher for Asians and Afro-Americans than it was for whites.
One needs to look no farther than China to understand the concerns of those fearing facial recognition technology. China is known for its advanced facial recognition systems that can track down people travelling across regions within minutes by monitoring the person’s movements using the millions of cameras installed. In what is being likened to a Black Mirrorish science fiction movie, the country is developing a social credit system, unveiled in 2014, which uses facial recognition, among other aspects to rate citizens. While some parts of it are already in place, the country is looking at 2020 as the year when it all comes together.
Through the credit system, citizens are rated based on daily activities which are promoted or looked down upon. Those that are frowned upon, such as jaywalking, smoking in no-smoking zones, effusing military service, reposting fake news online, will make a citizen’s credit score go down. This then affects their ability to perform functions such as buying airline or train tickets online, reduces internet speeds, bars them from enrolling in top schools or even from applying to certain jobs. Some are even being named and shamed.
Catching up to Big Brother
India is not far behind when it comes to surveillance camera coverage, as per a report by the British firm Comparitech, New Delhi, Chennai and Lucknow are among the top 50 cities in the world with the most CCTV cameras. New Delhi made it to the top 20 cities with nearly 3,00,000 CCTV cameras at 9.62 CCTV cameras per 1,000 people.
The Government is also planning to set up the world's largest automated facial recognition software (AFRS). In mid-2019, the National Crime Records Bureau (NCRB) issued a tender to build a national-level searchable platform for faces. As per reports, this database will be created using data from the Crime and Criminal Tracking Network and Systems (CCTNS), passports, the Interoperable Criminal Justice System, among others. NCRB has given March 27 as the new deadline for proposals for the AFRS.
This move prompted Internet Freedom Foundation, an advocacy body, to send a legal notice to the NCRB to halt and recall the invitation for bids for the implementation of AFRS. The group pointed at the arbitrariness with which such a move was being made, the lack of any legal framework and the absence of any safeguards against violation of privacy.
NCRB responded to the allegations, stating that there was no underlying statutory basis to the creation of the AFRS, by citing a CCTNS Cabinet note of 2009 where a similar system had been envisaged and had received Cabinet approval. The response also said that AFRS will use only CCTNS database which is secure, and will not rely on data from public CCTV cameras unless it is part of a crime scene. There also will not be any integration of AFRS with Aadhar, as per the response.
One cannot deny the advantages that facial recognition systems have for fighting crime such as ensuring faster processing, tracking criminals and terrorists and ensuring seamless integration. However, with the lack of a legal framework on the usage of such data, the system is highly susceptible to misuse. Hence, such systems will only be viable if the country were to have stronger data protection regulations to ensure that the technology is put to good use.