The tender published by the government agency has several vendor requirements, such as a portable facial recognition scanner for every 100 applicants.

You read it here first: The National Testing Agency (NTA), which conducts the Joint Entrance Examinations (JEE), National Eligibility cum Entrance Test (NEET), UGC-NET, and other key exams, will use facial recognition technology along with other biometric methods to screen candidates.

To understand how many people would be affected, JEE (Main) 2021 (February session) alone saw 6.52 lakh registrations, NEET 2020 was over 15 lakh registrations, and UGC-NET had 5 lakh registrations for the exams in September 2020. In fact, at one point in the tender – published by the NTA for selecting a company to install and implement this technology – “the total number of candidates for all exams in one year will likely be 50 lakh (approximately)”.

Why is it important? The Government of India has put great pressure on the introduction of facial recognition technology (FRT) in its various departments and institutions due to COVID-19. The rationale for this is that FRT would help meet social distancing standards and other COVID-19 guidelines; However, one has to wonder how, since FRT requires a person to take their mask off for review. Aside from the procedural aspects, the introduction of these technologies also comes at a time when India does not have robust laws regulating the use of such technologies that handle sensitive data. It is also important to highlight that by implementing FRT in the JEE and NEET exam centers, the NTA will handle sensitive facial data of many minors.

The main objective of this proposal, according to a call for tenders published by the NTA, is to control impersonation with the aim of biometric authentication and the presence marking service. Impersonation in entrance tests like JEE is nothing new and continues to this day. In 2020, Assam JEE Main 2020 Topper was arrested for allegedly using an impersonator to take an exam. In addition, the NTA recently used an AI algorithm to detect impersonation and identified 56 such candidates from the February cycle of the JEE Main 2021. Her pictures match some of the 20,000 top-ranked candidates for the 2019 and 2020 exams, one said report by Times of India.

This proposal, coupled with the recent Ministry of Health and Family Welfare (MoHFW) tender to introduce FRT for online exams conducted by the All India Institute of Medical Sciences (AIMS), brings much of the usual entrance exams under the field of the Face recognition. At the risk of repetition, the author has to reiterate that there is still no law in India that specifically regulates the use of personal data. India’s proposed 2019 Personal Data Protection Act, which addresses such concerns, could be brought to parliament at its winter session.

Real-time facial images for comparison with data from NTA

The NTS has provided various information about the functionality of face recognition:

  • Candidates are verified by capturing their iris, fingerprint and facial details in real time
  • The NTA provides center-related data such as role numbers, photos, name, exam date / shift etc. of all registered candidates. This data is used for comparison with the facial recognition data and other biometric data.
  • Candidate data from the NTA database at the time of review is obtained by scanning the QR code on a candidate’s admission card. Then the biometric verification takes place.
  • Two data centers will be set up. While one site serves as the main data center, the other is used for disaster recovery.
  • There will be hand-face recognition, fingerprint and iris scanners for all 100 candidates
  • Images are saved in GIF, PNG, JPEG, TIFF format and the image quality should be at least 60 pixels.

Live video surveillance in 600 examination centers

In addition to the verification by facial recognition before the start of the exam, the candidates are also subjected to video surveillance. According to the tender, the live material is to be transmitted to a control room in the NTA office in New Delhi. It must also be set up by the company that wins the contract for the tender. At least 95 percent of the live camera feed is always available in the control room during the test. It also said that video surveillance systems would be installed at least two days before the investigation. These are the likely locations where it will be installed –

In a place that covers all candidates sitting in the exam center without taking blind spots into account.

  • Entry / exit points of the examination center
  • stairs
  • Registration area
  • Hairdressing area
  • Central control room
  • server room
  • Center responsible room
  • Park
  • Driveway
  • Media connection

In addition to biometric and CCTV surveillance, the NTA has also proposed searching candidates using a portable metal detector (HHMD). As for CCTV surveillance, the NTA said the data will be kept for a period of 3 months. It did not mention any such clauses for the facial or other biometric data collected.

Experts say that the candidate’s explicit consent is a must

Shreya Suri, partner at IndusLaw, said that under the Information Technology Rules (Adequate Security Practices and Procedures and Sensitive Personal Data or Information) from 2011, biometric information such as “fingerprints”, “eye retina and iris”, “face patterns”, etc., are allowed Authentication purposes are considered “sensitive personal data”.

“Therefore, in order to avoid a conflict with the newly recognized right to privacy and existing laws, express prior consent (which is a crucial aspect in connection with such information under the existing and proposed law) in a clear and express from the data subjects whose data is collected and processed, and the NTA must also take into account the contractual capacity of the data subjects for this purpose (e.g. 18 years of age) – Shreya Suri, partner at IndusLAw

Similarly, Shweta Mohandas, Policy Officer at the Center for Internet and Society, pointed out that “The tender does not specifically prohibit the processing of the data if the data is kept by the bidder for 3 months, so the bidder may have time to review it To use data ”. to train their systems without asking for student consent again. “

Interestingly, the tender also transfers responsibility for maintaining the security of the data to the company chosen for the project. In this regard, Amay Jain, Associate at Victoriam Legalis said that “time will tell how effective these standards and best practices are in maintaining the strict privacy and confidentiality of the data to which the successful bidder in this tender has access”.

However, Siddharth Jain, co-founder of PSL Advocates and Solicitors, felt that there were adequate safeguards in place to prevent data breach mishaps. “Since the bidder in the tender document must comply with legal obligations and maintain strict confidentiality, including but not limited to the disclosure of the information to third parties, I believe that the privacy of the people is sufficiently protected and this exercise would be made easier.” Process of conducting exams, ”Jain said.

FRT in other educational institutions

Long before COVID-19 led to rampant adoption of such technologies, several educational institutions were testing facial recognition tools for purposes such as presence detection and wrongdoing verification.

  • Punyalok Ahilyadevi Holkar Solapur University in Solapur, Maharashtra: Better known as PAH Solapur University, the institution wanted to introduce a facial recognition-based attendance system for its 300 employees in April, according to the tender documents examined by MediaNama.
  • Rajiv Gandhi University of Health Sciences (RUGHS), Bengaluru: Last March, Deccan Herald reported that the RGUHS will use the face recognition of students in the academic year 2020-2021 to check treatment errors in medical and dental examinations. Deccan Herald had quoted the university vice chancellor as saying, “This is mainly to prevent wrongdoing like identity theft during exams.” He also said the Face ID will be linked to Aadhar.
  • Two schools in Chennai: Noisy Hindu, Tamil Nadu’s e-governance agency introduced an attendance system based on facial recognition in schools in Chennai. An official had made it clear to the newspaper that the cameras are not used for surveillance, but only remain active for the duration of the presence marking.
  • For admission to courses in Telangana: According to another report in the Hindu, the Telangana State Council of Higher Education introduced facial recognition systems that require students to upload a photo of themselves and the system to generate all the details. For this to work, a student had to download the Telangana government’s T App Folio app, upload a selfie, and provide details such as the Aadhar number and registered cell phone number.

also read

Do you have anything to add? Subscribe to MediaNama and post your comment