Competition for artificial intelligence supremacy is heating up on Asian-Western lines. Photo: iStock/Getty Images
AI is making fast strides in simulating and influencing human behavior. Photo: iStock

In Delhi, schools under the National Capital Territory’s jurisdiction now have closed-circuit cameras keeping watch around the clock.

Despite protests from civil liberty bodies, chief minister Arvind Kejriwal of the territory’s ruling Aam Aadmi Party defended the move. “There will be no privacy breach; children go to school for education, to learn discipline and become good citizens of the country,” he said. “They do not go there for anything private.”

A creeping surveillance net is spreading across the country as police forces and intelligence agencies increase their capacity for mass surveillance using technologies that, even if they’re fairly new, are readily available in the market.

Technologies ranging from simple surveillance cameras to artificial intelligence and facial recognition are deployed –  whether their targets like it or not. Meanwhile, the central government passes legislation narrowing citizens’ rights.

Facial recognition

On June 28 the National Crime Records Bureau issued a request for a proposal to build a “National Automated Facial Recognition System.” The proposal stated that its main objective is to modernize the police force and help in “information gathering, criminal identification, verification and its dissemination” among police forces across the country.

The NCRB is a federal body that maintains a statistical record of crimes across the country and also acts as a central repository of crime-related data. The move comes from the federal government, which offers to supply any data generated by the proposed system to the state governments.

In India, police powers are divided between the federal and state governments and that division is enshrined in the constitution. The bulk of the powers lies with the states, and the federal government steps in on rare occasions depending on the nature of the crime or when the state’s ability to carry out an impartial investigation is questioned.

This explains why a federal government body is keen to share the data under the facial recognition system with the states. “It will ensure minimum opposition from states ruled by parties that are opposed to the BJP, which it is in power,” a senior police official told Asia Times, referring to the Bharatiya Janata Party led by Prime Minister Narendra Modi.

International groups such as the UK’s Privacy International have been steadfastly critical of such moves by law enforcement and intelligence agencies. “While new technologies may be deployed under the guise of protecting democratic society,” the group said, “without adequate regulations and safeguards, those technologies can threaten democratic participation and dissent and thereby undermine democracy itself. This is not to say that new technologies should never be used; but the use of intrusive technologies requires transparent and robust oversight.”

While India inherited its law enforcement and intelligence mechanisms from colonial Britain, it has not been able to bring them under any parliamentary statute so far. In the UK, the Security Service, the Secret Intelligence Services and the General Communications Headquarters all have been defined by law since 1989 and also have a multi-layered oversight mechanism.

In India, none of the intelligence agencies has been created by an act passed by Parliament. As a result, there is no worthwhile oversight. That lack permits the government of the day to use its agencies as it sees fit.

The minimal oversight mechanisms in India are either too weak or simply non-existent.

According to information made available under the Right To Information Act a few years ago, the federal government puts nearly 10,000 telephones under surveillance annually. This means it clears nearly 300 requests for surveillance every day. The power to do so rests with the union home secretary.

“As a result, the time needed to apply any oversight on the request or its efficacy is nearly nonexistent,” a former intelligence chief told Asia Times. “The union home secretary has many other responsibilities, so paying attention to 300 orders a day is impossible,” this former chief said.

The NCRB plans to achieve its facial recognition system by maintaining a large database with photos and videos of faces through CCTV cameras and other sources like newspapers, raids and sketches. It claims that will “help to identify criminals, missing people, and unidentified dead bodies” and will help with crime prevention. This will be linked with other existing databases of Crime and Criminal Tracking Network & Systems, the Integrated Criminal Justice System, the Immigration, Visa and Foreigners Registration & Tracking and the national web portal on missing children.

However, it has repeatedly been found that facial recognition systems are discriminatory. A recent independent report concluded that four out of five people identified as criminal suspects by London Metropolitan Police were actually innocent. The police had an error rate of 81%, raising significant concerns on how they were using the technology.

Similarly, China used this system to restrict the rights of Uighurs. The US, which operates the largest facial recognition system, mines photos from driver’s licenses for facial recognition matches to target undocumented immigrants.

Targeted and mass surveillance

While the system may help in swift action, the downside of it is mass surveillance. In India, particularly, there already is an extensive database under the unique digital identification program – Aadhaar – which covers around 95% of the population. It contains iris imaging and biometric information along with fingerprint authentication. If face imaging is added, then the infrastructure of full surveillance will be complete.

Meanwhile, there have been various attempts at the state level to harness artificial technology. The Telangana state police force launched its own facial recognition system last year. Similarly, Punjab and Tamil Nadu state police forces use a face-recognition application called Facetagar.

The Union Ministry of Civil Aviation launched a DigiYatra’(Digital Travel) program, which aims to use facial recognition for the entry of passengers into the airport. The trial phase was rolled out at the Hyderabad airport on July 1 and saw 180 passengers register on the first day. While the system is advertised as ensuring that the passengers’ data automatically is erased once the flight departs, the major concern is the metadata that still remains in the system. Metadata are data that describe and give information about other data. So, the core issue remains: How is the data collected, stored and retained?

Old-tech threats via legislation

High-tech threats are not the only worry regarding civil liberties. Laws passed by the Parliament recently reinforce fears that civil liberties are in retreat. The latest move by the Modi government has empowered police to brand an individual a “terrorist” without going through a judicial process.

“To designate a person suspected to have terror links as a terrorist is necessary to root out terror,” Union home minister Amit Shah assured parliament earlier this week. “Anti-terror laws would not be misused, and used only to root out terrorism.”

However, those in India’s opposition parties were not convinced. The bill as proposed “violates Articles 14 and 21 and in my view, no one can be called terrorist merely on the basis of the feeling of the government or mere suspicion,” Asaduddin Owaisi, member of parliament from the Majlis-e-Ittehad-ul-Muslimeen party, argued. “One can be called a terrorist only after a judgment of a court of law. This is social ostracization.”

The speech by Owaisi, a barrister from the UK by training,  went viral and was widely hailed as he argued for a rollback of the government proposal.

Leave a comment