Author(s): Kirti Soni & Yashvardhan Singh Rajpurohit
Paper Details: Volume 4, Issue 3
Citation: IJLSSS 4(3) 40
Page No: 443 – 449
ABSTRACT
India’s criminal justice system depends on witnesses more than it is willing to admit. And yet, the state has done precious little to protect them. Witnesses are threatened, pressured, and in some cases killed while the system looks the other way. The result is predictable: witnesses turn hostile, the accused go free, and victims are left without justice. The Witness Protection Scheme of 2018, approved by the Supreme Court, was a step in the right direction but it is not enough. It has no legal teeth, it is unevenly implemented, and it leaves out too many cases. This paper looks at why witness protection has remained such a weak point in Indian criminal law, what courts and the Law Commission have said about it, what other countries have done, and why Parliament needs to stop delaying and pass an actual law.
I. INTRODUCTION
There is something deeply uncomfortable about the way India’s criminal justice system treats its witnesses. These are ordinary people, a shopkeeper who saw an assault, a neighbour who heard a threat, a bystander who witnessed a killing, who are called upon to do what most of us would consider a basic civic duty: tell the truth in court. But in India, telling the truth in court can cost you your livelihood, your safety, and sometimes your life. And the state, which summons these people and depends entirely on their testimony to make its case, has no serious mechanism to protect them.
This is not a new problem. It has been studied by the Law Commission, lamented by the Supreme Court, discussed in parliamentary committees, and written about extensively in legal journals. Yet India still does not have a dedicated statute on witness protection. What it has is an executive scheme, approved by the Supreme Court in 2018, that is not backed by any law, carries no penalties for non-compliance, and has been implemented in a half-hearted, piecemeal manner across most states.
The consequences of this failure are visible in almost every major criminal trial. Witnesses retract. Cases collapse. The accused, often better resourced and better connected than the people testifying against them, walk free. This paper tries to understand why this keeps happening, what the law currently offers, and what a genuine solution would look like.
II. WHY WITNESSES MATTER AND WHY THEY ARE SUFFERING
India follows an adversarial system of criminal justice. Unlike inquisitorial systems, where judges take an active role in gathering facts, our system depends on the parties, that is the prosecution and the defence, to put evidence before the court. The Bharatiya Sakshya Adhiniyam, 2023 (which replaced the Indian Evidence Act) continues this tradition. Oral evidence, meaning what witnesses say from the stand, is still the primary mode of proof. Documents and forensic material need to be proved through witnesses. Even a confession is meaningless unless the circumstances around it are established, usually through the testimony of the investigating officer.
What this means in practice is that the moment a witness stops cooperating, the prosecution’s case begins to unravel. This is not a theoretical concern. In the Best Bakery case arising from the 2002 Gujarat riots, witnesses who had given statements identifying the accused completely retracted their testimony when the trial was held in Gujarat. The Supreme Court had to intervene and order a retrial in Maharashtra. In the Jessica Lal murder case, over a hundred witnesses turned hostile. Cases of custodial torture, dowry deaths, and sexual assault routinely fall apart because the people who were there are no longer willing or able to say what they saw.
The reasons are not hard to understand. Physical intimidation is the most obvious one: witnesses receive threats, their families are approached, and in some cases they are attacked. But economic pressure is just as powerful and far less discussed. Every time a witness has to come to court, sometimes dozens of times over years of adjournments, they lose a day’s wages. The allowances paid under the Bharatiya Nagarik Suraksha Sanhita are inadequate and rarely paid on time. For a daily-wage labourer or a small trader, prolonged engagement with the criminal justice system is simply not something they can afford.
And then there is the sheer length of trials. India’s courts carry an enormous backlog. Over four crore cases were pending in district courts as of early 2024 according to the National Judicial Data Grid. Many criminal trials run for a decade or more. The Law Commission’s 198th Report on Witness Identity Protection and Witness Protection Programmes (2006) noted that in several states, more than half of all witnesses in criminal trials eventually turn hostile. That is not a figure that reflects individual dishonesty. It reflects a system that has failed to make truth-telling safe.
III. WHAT THE LAW CURRENTLY OFFERS AND WHERE IT FALLS SHORT
The most significant development in this area is the Witness Protection Scheme, 2018, which the Supreme Court approved in Mahender Chawla v. Union of India (2018) 16 SCC 524 and directed all states and Union Territories to implement immediately. The scheme divides witnesses into three threat categories: Category A covers a threat to life extending beyond the trial, Category B covers threats to safety, reputation, or property, and Category C covers a moderate threat. Based on the category, protective measures range from police escort and in-camera proceedings to relocation and identity change. Each state is required to constitute a Witness Protection Cell and set up a Witness Protection Fund.
On paper, this looks reasonable. In practice, it has major problems. The most fundamental one is that the scheme is not a law. It was approved by a court order. A state that ignores it cannot be prosecuted under any statute. It can only be dragged into contempt proceedings before the Supreme Court, which is both slow and politically contentious. Several states have simply not constituted the mandated cells or created the fund. Others have done so in name but not in any meaningful sense.
The scheme also covers a very narrow category of cases, mainly organised crime, terrorism, and offences punishable with death or life imprisonment. This leaves out dowry deaths, acid attacks, custodial torture, and many cases of communal violence, where witnesses are just as vulnerable. There is no provision for economic support either. No compensation for lost wages, no legal aid, no help with rehabilitation if a witness has to be relocated. Relocation without financial support is meaningless for most ordinary people, who cannot simply leave their jobs and communities behind.
Outside the scheme, there are scattered provisions across different statutes. The BNS protects the identity of rape victims. The POCSO Act, 2012 allows child witnesses to testify via video-link and provides for support persons. The Prevention of Money Laundering Act has in-camera hearing provisions. These are useful, but they address specific situations rather than providing a general framework. A witness in a murder case involving an ordinary accused has very little protection under any of these.
Section 228 of the Bharatiya Nyaya Sanhita, 2023 (previously Section 195A IPC) does make witness tampering a criminal offence. But it requires the witness who has been threatened to file the complaint. That means the very person who is least likely to come forward when the threat has actually worked must be the one to initiate action. This is a structural absurdity that has been noted by courts and commentators alike, but has not been fixed.
IV. WHAT COURTS AND THE LAW COMMISSION HAVE SAID
The Supreme Court has been clear and consistent about the seriousness of this problem. In the Best Bakery case, Zahira Habibulla Sheikh v. State of Gujarat (2004) 4 SCC 158, Justice Arijit Pasayat observed that witnesses in India are treated like “bonded labourers of the system,” required to serve without support or protection. The Court held that the state has a positive obligation to ensure that witnesses can testify safely, and linked this obligation to the right to a fair trial.
In Mahender Chawla, the Court went further and grounded witness protection in Article 21 of the Constitution. Justice K.M. Joseph, in his concurring opinion, explicitly noted that the Witness Protection Scheme must eventually be replaced by legislation, because an executive scheme cannot provide the certainty or the enforceability that the situation demands. This observation, coming from the Supreme Court itself, is about as clear an instruction to Parliament as one can imagine. Parliament has not responded.
The Law Commission’s 198th Report, submitted in 2006, is the most thorough study of this issue from an official body. After examining witness protection laws in the United States, United Kingdom, Canada, and South Africa, the Commission proposed a draft Witness Protection Bill for India. It recommended a National Witness Protection Authority with statutory powers, a time-bound threat-assessment process, economic assistance for witnesses, criminal sanctions for witness tampering with a reversed burden of proof, and provisions for identity concealment, relocation, and in-camera hearings across all serious criminal cases. The report was submitted nearly twenty years ago. No bill has been introduced in Parliament.
V. WHAT OTHER COUNTRIES HAVE DONE
It is worth briefly looking at how other countries have handled this, because the comparison is instructive. Not as a reason to copy foreign law blindly, but to show that the problem is solvable.
The United States’ Witness Security Program (WITSEC) was established under the Organized Crime Control Act, 1970. It is run by the US Marshals Service and provides protected witnesses with completely new identities, new homes, vocational training, medical care, and a monthly allowance until they become self-sufficient. It is a federal statute, independently funded, and administered by a dedicated professional agency. The programme has protected over 19,000 witnesses and their families. Not one witness who followed the programme’s rules has been killed.
South Africa’s Witness Protection Act of 1998 is probably the most relevant model for India. South Africa has a colonial legal history similar to ours, comparable levels of economic inequality, and serious challenges with organised crime. The Act creates a statutory Director of Witness Protection with the power to provide relocation, new identity documents, and financial support to witnesses in any serious criminal case, not just terrorism or organised crime. Witnesses enter into a formal agreement with the state and can challenge decisions about the nature of their protection. There are oversight mechanisms and accountability structures built in.
The United Kingdom’s Coroners and Justice Act, 2009 allows witnesses to remain anonymous in court where disclosure of their identity would endanger their life or safety. The decision requires judicial approval, is subject to review, and is balanced against the accused’s right to confront witnesses. This is a carefully calibrated model that takes both witness safety and fair trial rights seriously.
India has the legal and administrative capacity to enact something similar. The POCSO Act demonstrates that central legislation can prescribe minimum standards for witness-related procedures that states then implement. There is no constitutional barrier to Parliament legislating on witness protection. Criminal law and criminal procedure are concurrent subjects under the Seventh Schedule.
VI. WHAT NEEDS TO CHANGE
The argument of this paper comes down to something simple: India needs an actual law on witness protection, not just a scheme. The Witness Protection Scheme of 2018 was better than nothing, but it was never going to be sufficient on its own, and the Supreme Court said as much when it approved it.
A proper Witness Protection Act should do several things. It should create a National Witness Protection Authority, a statutory body with real powers, adequate funding, and independence from local law enforcement and political pressure. It should apply to all serious criminal cases, not just terrorism and capital offences. It should provide economic support in the form of wage compensation for court appearances, legal aid for witnesses who need it, and rehabilitation assistance for those who are relocated. It should make anonymity, in-camera proceedings, and video-link testimony available as default options wherever a credible threat exists.
The law should also fix the witness tampering provision. The current position, which requires the threatened witness to file the complaint, is unworkable. A witness who is being threatened effectively enough to change their testimony is not going to walk into a police station and report it. The law should allow the prosecution or an independent authority to initiate proceedings on the basis of evidence of tampering, and should place the burden on the accused to disprove involvement once a prima facie case is established.
Implementation matters just as much as the text of the law. India has a long history of good legislation that has failed at the implementation stage. Any witness protection statute must include mandatory training for police and prosecutors, specific timelines for threat assessment and decision-making, independent audit of state-level implementation, and a right for witnesses to seek judicial review if they believe the state has not fulfilled its obligations toward them.
VII. CONCLUSION
It is difficult to look at India’s record on witness protection without feeling that something has gone seriously wrong, not in the law on paper but in the priority the state has given to this issue. The people who provide testimony in criminal courts are doing something important and risky. They deserve more than a scheme that exists mostly on paper.
The Law Commission recommended a proper law in 2006. The Supreme Court linked witness protection to the fundamental right to life in 2018 and strongly suggested that legislation was necessary. Multiple parliamentary committees have flagged the issue. The political will to act, however, has been absent.
This cannot continue. Every criminal trial in which witnesses are intimidated into hostility is not just a miscarriage of justice in that individual case. It is a signal to every potential witness in every future case that the state cannot protect them. That signal has consequences. It makes people less willing to come forward, less willing to cooperate with the police, and less willing to trust a system that asks for their help but offers nothing in return.
A witness protection law will not solve everything. But it is a necessary starting point. Parliament needs to act, and it needs to act with more seriousness than it has shown so far.
SELECT BIBLIOGRAPHY
CASES
Mahender Chawla v. Union of India (2018) 16 SCC 524.
Zahira Habibulla Sheikh v. State of Gujarat (Best Bakery Case) (2004) 4 SCC 158.
State of U.P. v. Ramesh Prasad Misra (1996) 10 SCC 360.
LEGISLATION AND SCHEMES
Bharatiya Nyaya Sanhita, 2023; Bharatiya Sakshya Adhiniyam, 2023; Bharatiya Nagarik Suraksha Sanhita, 2023.
Protection of Children from Sexual Offences Act, 2012.
Witness Protection Scheme, 2018 (Supreme Court of India).
South Africa Witness Protection Act 112 of 1998.
UK Coroners and Justice Act, 2009; US Organized Crime Control Act, 1970.
REPORTS AND BOOKS
Law Commission of India, 198th Report on Witness Identity Protection and Witness Protection Programmes (2006).
R.V. Kelkar, Criminal Procedure (6th, Eastern Book Company 2014).
Ratanlal and Dhirajlal, The Law of Evidence (27th, LexisNexis 2022).
National Judicial Data Grid, Pendency Statistics (2024).
PUCL, ‘Witness Protection in India: A Report’ (2019).