Springing a surprise, the Union Minister of State for Home Affairs, Ajay Mishra Teni, on Monday introduced the Criminal Procedure (Identification) Bill 2022. The Bill was neither put up for pre-legislative consultation nor indicated in the session’s legislative agenda in Parliament. Seemingly technical, it is a legislative proposal that undermines the privacy of not only persons convicted of crime but also every ordinary Indian citizen as it proposes replacing a law that is over a 100 years old.
What needs scrutiny
Let us first understand why it is being introduced, and what it intends to achieve. The Bill aims to replace the Identification of Prisoners Act 1920 that has been in need of amendment for several decades. Back in the 1980s, the Law Commission of India (in its 87th Report) and the Supreme Court of India in a judgment titled State of U.P. vs Ram Babu Misra had nearly simultaneously suggested the need to amend the statute. The criticism and the need for amendment was predominantly in respect of the limited definition of ‘measurements’ as under that Act. It seems that this is one of the primary issues that the proposed legislation is designed to resolve.
In this regard, it might be unexceptional, being an expression of long-held views within the legal establishment. However, the devil is in the details, with three expansions in the power of state surveillance (in the name of criminal reforms) that merit further scrutiny.
First, the definition of measurements is not restricted to taking measurements, but also their “analysis”, when the definition now states “iris and retina scan, physical, biological samples and their analysis, behavio[u]ral attributes including signatures….” This definition is nebulous and vague. It goes beyond the scope of a law which is only designed for taking measurements and could result in indirectly conferring legislative backing for techniques which may involve the collection of data from other sources. For instance, using facial recognition technology where measurements of persons as under this law are compared with samples taken from the general public.
At present there are extensive facial recognition technology programmes for “smart policing” that are deployed all across the country. For instance, the Delhi police use facial recognition technology originally acquired for identification of missing children in 2018 to also screen for “habitual offenders”. Similarly, the Tamil Nadu police deploy facial recognition systems which are integrated with State- and national-level databases including CCTV footage. Such experimental technologies cause mass surveillance and are prone to bias, impacting the fundamental rights of the most vulnerable in India.
Data capture and ‘choice’
The second area of the expansion of surveillance concerns from whom such “measurements” can be gathered. The existing law permits data capture by police and prison officers either from persons convicted or persons arrested for commission of offences punishable with a minimum of one year’s imprisonment. Parallel powers are granted to judges, who can order any person to give measurements where it is in aid of investigation. While the judicial power is left undisturbed, it is the powers of the police and prison officials that are being widened. The law removes the existing — albeit minimal — limitation on persons whose measurements could be taken. It is poised to be expanded to all persons who are placed under arrest in a case. This is a truly breathtaking spectrum, including petty crime such as violating a prohibitory order for not wearing a mask, jaywalking or a traffic violation.
Here, the proposed Bill also contains muddied language stating that a person, “may not be obliged to allow taking of his biological samples”. This, on its surface, offers a choice to a person to refuse. However the words “may not be obliged” may also be read to offer discretion onto a police officer to confer such a choice. In any instance the exercise of such “choice” is presumed in law, it may not be truly voluntary, given the absence of wider accountability reforms in which existing policing practices are coercive.
Even if these objections are disregarded, the “choice”, if any, is limited only to, “biological samples” from the wider data points captured within what constitutes, “measurements”. For instance, “iris and retina scan” is mentioned separately to, “biological samples”, and hence a person arrested under any crime or preventive detention law if desired by the police will be required to scan their eyes.
Storage of data
The third area of concern is the database of the “measurements” which are gathered. The National Crime Records Bureau (NCRB) shall for a period of 75 years from the date of collection maintain a digital record, “in the interest of prevention, detection, investigation and prosecution of any offense”. As pointed out by Prof. Aparna Chandra (an associate professor of law) on Twitter, “How will these records be used for preventing crime except through surveillance?” This becomes clear when the provision permits the NCRB to, “share and disseminate such records with any law enforcement agency, in such manner as may be prescribed”.
It is important to consider that the NCRB already operates a centralised database, namely the Crime and Criminal Tracking Network & Systems (CCTNS), without any clear legislative framework. The interaction between the proposed law and CCTNS is not clearly defined though likely, given the powers conferred under for digital records go to the same government department.
The existence of such legislative power with a technical framework may permit multiple mirror copies and parallel databases of the “measurements” being stored with law enforcement, beyond a State Police department which will be prosecuting the crime and the NCRB which will store all records centrally. For instance, in response to a Standing Committee of Parliament on police modernisation, Rajasthan has stated that it maintains a ‘RajCop Application’ that integrates with “analytics capabilities in real-time with multiple data sources (inter-department and intra-department)”. Similarly, Punjab has said that the “PAIS (Punjab Artificial Intelligence System) App uses machine learning, deep learning, visual search, and face recognition for the identification of criminals to assist police personnel. This app helps in storing and carrying information about criminals”. Hence, multiple copies of “measurements” will be used by State government policing departments for various purposes and with experimental technologies. This also takes away the illusionary benefit of deletion which occurs on acquittal and will suffer from weak enforcement due to the absence of a data protection law.
In sum, once a person enters their “measurements” within the system, they stay there for life given the average life expectancy in India which hovers around 70 years is less than the retention period. The end result is a sprawling database in which innocent persons are treated as persons of interest for most of their natural lives.
While the impact on persons with privilege may be minimal, the masses — many of whom lack social and economic power in Indian society — may face harsher law enforcement. This becomes clear from the primary research-based article, “Settled Habits, New Tricks”, by Ameya Bokil, Nikita Sonavane and Srujana Bej from the Criminal Justice and Police Accountability Project (the other writers include Avaneendra Khare and Vaishali Janarthanan). They pointed to the caste bias against the Pardhi Adivasi community which was at one time designated as a criminal tribe. In this context they state, “In reality since these databases are fed by the police’s centuries-long caste-based system of preventive surveillance and predictive policing (which has already determined who is a criminal and what crimes habitual criminals commit repeatedly), there is no possibility of objectivity or lack of caste bias. The CCTNS only adds a technological veneer to a caste-based policing model....” It is foreseeable that if the proposed ambit of “measurements” is expanded and then put in a database, it will likely also target the Pardhis.
Onus is on government
Injuries to privacy are not mere academic debates and cause real, physical and mental consequences for people. To protect individual autonomy and fulfil our constitutional promises, the Supreme Court of India pronounced the Justice K.S. Puttaswamy judgment, reaffirming its status as a fundamental right. The responsibility to protect it falls to each organ of the government, including the legislature and the union executive. For India to fulfil its claims of being a constitutional democracy, rather than a mere electoral democracy, it will have to be better rather than regressing even from the Identification of Prisoners Act passed by a colonial regime.
Apar Gupta is a lawyer and the Executive Director of the Internet Freedom Foundation. Abhinav Sekhri is an advocate practising in Delhi