When the Police Version Becomes the Evidence: What the Allahabad High Court’s Encounter Cases Reveal About State Power

There is a peculiar challenge about police encounter cases rarely discussed. The police are usually the first ones to narrate what has happened, the custodians of the initial evidence, and the source from  which an investigation starts. When a court takes up such a matter, the State has already been able to present its version of events.

Its version may well be true; it may also only contain the truth in parts; in a serious case, it may even be wholly false.

The legal issue before the court thus goes beyond whether the police fired correctly; it examines whether their version of events can withstand scrutiny.

That is where Allahabad High Court’s recent pronouncements on police encounter cases become relevant.

In January 2026, in Raju alias Rajkumar v. State of U.P., the Court took up the matter concerning persons who were grievously injured in encounters. It was found that the required safeguards under People’s Union for Civil Liberties v. State of Maharashtra were not put in place and observed that those protections applied to grievous injuries. It further questioned the growing practice of police using shots to the legs and calling it an encounter. (Raju alias Rajkumar v. State of U.P. paras 6-13)

A few months later, in Chotkau alias Allauddin v. State of U.P., while examining the similar matter, the Lucknow Bench went one step back. Not only did the Court scrutinize whether the safeguards were installed, it examined the plausibility of the police version of events. It asked whether the number of personnel, the vehicle they traveled in, the division of teams, the firing, the injuries, the investigation, and even the rewards to the police personnel were consistent with the narrative. Having found it to be prima facie false, it directed a CBI inquiry. (Chotkau alias Allauddin v. State of U.P., Criminal Revision No. 835 of 2026, order dated 13 August 2026, paras 32-54.)

Taken together, both these judgments highlight an issue far beyond the so-called “encounter culture”.

They show that the constitutional courts are beginning to take a step back from the encounter as a closed event.
The police version of events, which has long been the default in such matters, is now being re-examined from scratch.
That is where the real value of Allahabad High Court’s recent rulings lies.

An encounter is an allegation of force.

The term itself covers many possibilities. At the most basic level, it suggests that both the police and an armed accused opened with fire and one of them was injured. But that is not necessarily the case. The phrase “police encounter” is not a safeguard of police conduct but rather an open accusation that the police used force – not necessarily lethal force – to apprehend an accused. The Supreme Court acknowledged this issue in People’s Union for Civil Liberties v. State of Maharashtra, a historic judgment concerning police encounters. (PUCL v. State of Maharashtra, (2014) 10 SCC 635, paras 31.1-31.16, 32-33.)

The Court examined the matter after hearing allegations regarding numerous encounters in Maharashtra. It concluded that the mere fact that a death occurred in an encounter could not be sufficient grounds for a police investigation into such an occurrence. Instead, a set of guidelines was established to ensure that such an incident was thoroughly investigated. Some of the most important measures include registering a First Information Report and conducting an independent CID or police inquiry, as well as preserving forensic evidence, with magisterial and medical evidence, and ballistics tests. Simultaneously, it was ruled that no immediate out-of-turn promotion or gallantry awards could be given to police personnel, unless it was proven that they displayed extraordinary courage in an encounter. The Court further held that the requirements would apply, as far as practicable, even if the accused suffered grievous injuries. (PUCL v. State of Maharashtra, paras 31.1-31.16, 32-33.)
It is worth noting that the various safeguards seem rather tedious.

That is because they are.

Each of the measures enumerated by the Supreme Court addresses a specific vulnerability. An investigation led by senior officers not present on the day of the incident allows for more objective oversight. Forensic examination should be able to cross-examine police statements with concrete evidence. Medical evidence, in turn, could corroborate or refute the police narrative of events. A magisterial inquiry provides an additional level of control, and restrictions on rewards help prevent rash celebrations and promotions before all facts have been established. These individual steps, therefore, fit within a broader principle.

They are supposed to ensure that the State does not get away with claiming that an encounter happened merely because it did.

That is why Allahabad High Court’s recent judgments are so critical.

Raju: the Court draws a distinction between defense and punishment

The January 2026 judgment in Raju alias Raj Kumar is significant for multiple reasons. However, it is notable primarily for having declined to accept the use of firearms to effect an arrest in general and to shoot at the legs as a means of punishment. The Court observed that the applicant before it had suffered grievous injuries as a result of an alleged encounter. It was further noted that the police had failed to record the injured person’s statement before a Magistrate or Medical Officer. Additionally, an investigation was not conducted by an officer senior to the head of the police party. It was found that the police had violated the Supreme Court’s safeguards for encounters. (Raju alias Rajkumar v. State of U.P., paras 6-10.)

The Court then made an even more intriguing observation. It said that it had seen a number of cases, even those involving minor offenses like theft, where firing was used in encounters. The police had legitimate grounds to use private defense, but it distinguished between essential force and punishment through force. (Raju alias Rajkumar v. State of U.P., paras 7, 11-12.)

That, as well, is a critical observation. A police officer can use force when the law permits them to do so. But the justification for using it must always come from the need to protect or defend, and not from a sense of entitlement. The Court ruled that punishment falls under the jurisdiction of the judiciary and not the police. It further clarified that the fact that the injury was to non-critical body parts did not exonerate the police of the responsibility to follow due process. (Raju alias Rajkumar v. State of U.P., paras 11-13.)

This seems to be the most essential takeaway from the January ruling.

There is a tendency to believe that a “leg shot” is a convenient middle ground between arrest and death.

But the Court warned against viewing it as a lesser evil. If the police choose to shoot to maim as a form of punishment, that does not make it any less of a violation. Its reasoning may well extend to fatal encounters, for the Court notes that the lack of police injuries creates questions about the necessity and proportionality of force used.
(Raju alias Rajkumar v. State of U.P., paras 18-20.)
That is not to say that the physical injuries to the police are not a consideration. But the Court warned against the police using that as a justification for using excessive or disproportionate force. The physical evidence must always be examined with a view toward determining whether the force used was appropriate to the situation.

Chotkau: When The Court Stopped Reading The FIR As Gospel

The Chotkau order goes considerably further in scrutinising evidence than the FIR.

The background to the case is unusual. Chotkau had previously been convicted of rape and murder and had received the death penalty. The Supreme Court acquitted him in September 2022. Doing so, it criticised the assessment of evidence by the courts below and drew attention to some serious contradictions, insisting that even in a heinous case, evidence will have to withstand careful scrutiny. It also noted that Chotkau was unable to afford legal representation in the Sessions Court and had eventually been represented by an Amicus Curiae. (Chhotkau @ Allahuddin v. State of U.P., (2023) 6 SCC 742, paras 85-88.)

This history became relevant when the same person was later subjected to the alleged encounter in May 2025.

The police version was elaborate. Thirteen police personnel were allegedly travelling in a single government vehicle while searching for Chotkau, who was said to be attempting to escape towards Nepal. A SWAT team was then called as reinforcement, taking the total number of police personnel to 23. The teams allegedly took positions near a bridge, intercepted Chotkau on an e-rickshaw and surrounded him. According to the FIR, he opened fire at the police, threatened to kill them and began loading his weapon again. The SHO then fired two rounds, both of which struck Chotkau in his legs. (Chotkau, paras 6-16.)

The Court did something that is particularly significant for the future of encounter litigation.

The Court tested the mechanics of the story.

The FIR stated that thirteen personnel had been moving in one government vehicle and were subsequently divided into three teams. The Court questioned whether such a vehicle could physically accommodate thirteen police personnel and how those three teams could move in separate directions without separate vehicles being accounted for. (Chotkau, paras 9-13.)

Then came the larger question.

How, the Court essentially asked, could 23 police personnel, including a ten-member SWAT team, fail to stop a person travelling on an e-rickshaw?

The issue was not whether the judge personally believed the police officer.

The issue was whether the sequence described in the FIR was internally plausible.

That is a fundamentally evidentiary mode of judicial review.

The same approach continued when the Court examined the shooting itself. The SHO told the Court that he had fired from approximately 15 metres away at night and that it was a moonlit night. The Court questioned whether he could accurately aim and fire a 9 mm service pistol under those circumstances while reacting merely to the sound of a weapon being loaded. (Chotkau, paras 43-48.)

The Court also examined the medico-legal report. It recorded that the wounds were of particular sizes, with no clear indication of entry and exit wounds. When the SHO was asked how a 9 mm bullet could produce the smaller wound described in the medical record, his explanation was that the tip of the bullet was narrower. The Court found that explanation unconvincing. (Chotkau, paras 47-48.)
None of these observations, taken alone, necessarily proves that an encounter was staged.

Taken together, however, they can justify the much more limited conclusion that the official version requires independent investigation.

Exactly that conclusion the Court reached.

The investigation itself became part of the problem

Perhaps the most revealing feature of Chotkau is that the Court did not stop at questioning the shooting. It examined what happened afterwards.

The case had initially been investigated by a subordinate of the SHO who had lodged the encounter FIR. The investigation was later transferred to another police station, but the charge sheet was submitted by the SHO of Ikauna Police Station, the same police station whose officer had been the complainant in the encounter case. (Chotkau, paras 49-53.)

This is where the logic of PUCL becomes especially relevant.

The Supreme Court did not require an independent investigation merely as bureaucratic formality. It recognised an institutional conflict of interest. The officer whose conduct forms part of the incident cannot be allowed to become the person through whom the incident is effectively validated.

The Supreme Court returned to this concern in Arif Md. Yeasin Jwadder v. State of Assam, 2025 INSC 785. While considering allegations concerning multiple police encounters, the Court stated that allegations involving police excesses could not fairly be investigated by the State Police where the allegations themselves concerned police personnel. The Court stressed that an inquiry into such allegations must be independent and insulated from institutional bias. (Arif Md. Yeasin Jwadder, paras 47-52.)

Importantly, the Supreme Court did not simply assume that every encounter in Assam was illegal. It expressly recognised that some encounters may ultimately turn out to have been necessary and legally justified. The point was that this determination had to come from a fair and impartial investigation rather than assumption. (Arif Md. Yeasin Jwadder, paras 23, 32-38.)

That distinction is important in Uttar Pradesh as well.

Judicial scrutiny of encounters does not require an anti-police position. It requires an anti-immunity position.

The reward question is more serious than it looks

The Chotkau order also recorded that all 23 police personnel involved in the alleged encounter had been rewarded for the operation. The Court noted that the alleged “good work” included extracting a confession concerning an offence in respect of which Chotkau had already been acquitted by the Supreme Court. (Chotkau, para 44.)

This raises an uncomfortable institutional question.
What happens to the neutrality of an investigation when the State has already rewarded the officers whose conduct must subsequently be examined?

The answer is partly contained in PUCL. Paragraph 31.15 specifically provides that out-of-turn promotions and instant gallantry rewards should not be given soon after an encounter. Such rewards should follow only after the gallantry of the officers has been established beyond doubt. (PUCL, para 31.15.)

The rule makes practical sense. A reward is not merely a monetary or professional benefit. It is also an institutional signal that the conduct was commendable.
Once that signal has been sent, subsequent scrutiny is inevitably harder.

This is why encounter accountability cannot be treated only as a question of criminal law. It is also a question of administrative incentives. A policing system should reward effective law enforcement, but it should be extremely careful not to create a professional incentive structure in which firearm use itself becomes evidence of successful policing.

The Problematic Part About The Judgement

The Court’s strongest observations came from the broader circumstances surrounding the alleged incident. The FIR stated that Chotkau had fired at the police but that none of the police personnel had been injured. It further stated that the police team sought additional reinforcement despite already having 23 personnel, many of them armed. The Court found it difficult to reconcile this with the fact that the alleged accused had only a country-made pistol and two cartridges, one of which was allegedly fired and the other recovered.  (paras 40-42.)

The Court also noticed that the original FIR concerning the kidnapping had been registered within roughly an hour of the alleged incident and did not contain the later allegations concerning rape and bleeding injuries. Those allegations subsequently appeared in the encounter narrative. The Court therefore stated that, prima facie, a false confession appeared to have been recorded. (paras 30-31, 38.)

Again, these are not final findings after trial but rather judicial reasons for demanding investigation. That distinction should remain central to any responsible discussion of these cases.

The High Court did not convict the police officers. It did not declare every encounter in Uttar Pradesh fabricated. It found sufficient discrepancies to conclude that the allegations could not safely remain within the ordinary investigative structure and therefore directed a CBI inquiry. (paras 52-56.). The difference between those propositions matters.

The Larger Constitutional Issue Here

The discussion about encounter deaths tends to focus on rights of criminals. That is the wrong way to think of it. The real issue is whether the State can take justice into its own hands, or whether it must always leave it to the courts.

Article 21 does not apply only to those whom the police suspect of grave crimes; the Allahabad High Court specifically refers to encounter guidelines as being cut out “in the larger interest of protecting life and limb and the right to dignity of individuals as enshrined in Article 21 and the primacy of the law as enshrined in Article 141 of the Constitution.” (Raju, paras 10-13)

The Supreme Court in Om Prakash v. State of Jharkhand has occasion to discuss the difference between legitimate police protection measures against anticipated accusations and police killings because the accused turns out to be a dreaded criminal. It states, “It is neither the function nor the duty of the police to kill the accused persons merely because they are dreaded criminals. On the contrary, such persons ought to be arrested and brought to trial.” (Om Prakash v. State of Jharkhand, (2012) 12 SCC 72, para 38.)

The principle that emerges, as we discuss, is that while the police may have the duty to use force, that does not make them judges of punishment.

The distinction is even sharper when it comes to an accused who is injured. A man shot in the leg is not necessarily subjected to less severe punishment.

But if the shot was fired unnecessarily, it does not become constitutionally acceptable merely because the victim survives.

What the Allahabad High Court is actually changing

The most interesting aspect of this development is not really that the judiciary has begun to suspect the police. It is that the Court has become willing to scrutinize the very foundations of the encounter narrative as soon as it emerges. That can be momentous.

Imagine the difference between a situation where the police say: “The accused fired at us first, so we returned fire and injured him. We recovered a weapon, and an FIR has been registered.

All of this is normal.

Now imagine a world where every claim in that narrative is subject to challenge.

Was there a weapon?

Do the ballistics reports match the injury?

Where exactly were the police personnel?

Was the range consistent with the injury?

Do the post-mortem or medical reports correspond to the FIR?

Were the police personnel in the vehicles as stated, and in the positions as mentioned?

Does the medical report substantiate the police narrative?

Was the investigation independent?

Were rewards announced before investigating the encounter?

Were there any additions to the narrative after the FIR?

The second world is a more thoroughgoing republic, if you will. But it also has the benefit of being safer in a constitutional democracy.

A good way to think of it is that the Allahabad High Court is now focusing more on the FIR as constituting the beginning of the encounter as evidence. Which may well be the most interesting part of this development.

Lawyer’s Perspective

Sudhanshu Khetrapal, Advocate

Speaking to LCI, Adv Khetrapal stated that “The real issue in these ‘encounter’ cases is not whether police should be able to take people down forcefully. They should, where the law allows. The issue is that their narrative must not be presumed to be true simply because they are the ones that took action. The recent turn of events in Allahabad are significant from this perspective as courts begin to look not merely at what happened but at the supporting evidence, the medical reports, and the independence of investigation. This is not anti-police sentiment. This is a basic standard of accountability that kicks in when the State uses its most coercive methods.”

Why this matters outside of Uttar Pradesh

The problem with the encounter debate is that it becomes very polarizing very quickly.

One side begins to believe that every encounter is a sign of a lawless State. The other side begins to believe that every encounter is a sign of a competent police force. Neither view is particularly useful to a court of law.
Courts have to evaluate each case on its own merits. That means that the criminal record of the accused cannot be ignored, but that also means that the popularity of the police officer cannot override the evidence before the court.

The Supreme Court in Arif Md. Yeasin Jwadder has occasion to discuss this question as well while dismissing several encounter cases. It emphasizes the need to distinguish each case on its own merits even while acknowledging that some encounters may genuinely be beyond the control of the police. (Arif Md. Yeasin Jwadder, paras 32-38)

That is the most important lesson that the Allahabad High Court’s orders are attempting to drive home.

There is no obligation for police personnel to stand inert in the face of a genuine threat. There is, however, a constitutional obligation for the State to demonstrate that the police narrative is credible.

That is not always easy if there is inconsistency within the narrative or if the injured person’s injuries do not match the claimed script or if questions are raised about the independence of investigation, the credibility of forensic evidence or the timing of rewards to police personnel. That is why we are seeing so many detailed questions being raised in Raju and Chotkau’s cases in particular.

Conclusion

The phrase “fake encounter” often creates the impression that the central question is whether the encounter was genuine or staged.

The more important legal question comes before that.
Can the State prove, through an independent process, that its use of force was lawful?

The Allahabad High Court’s recent cases suggest that this question can no longer be answered simply by reproducing the police version in an FIR.

In Raju, the Court focused on the failure to follow mandatory safeguards and warned against the use of firearm injuries as a substitute for punishment. In Chotkau, it examined the internal logic of the police account and found enough inconsistencies to require a CBI investigation. The Supreme Court’s decisions in PUCL and Arif Md. Yeasin Jwadder provide the larger framework: encounters are not outside the law, police officers are not presumed guilty merely because force was used, but neither is their account entitled to immunity from independent scrutiny.

That is an important distinction.

A functioning criminal justice system does not ask the police to surrender their authority.

It asks them to exercise that authority within law.

The ultimate test of policing in a constitutional State, therefore, is not how effectively an accused person can be neutralised. It is whether, after the force has been used, the State can place the facts before an independent authority and withstand the questions that follow.

The Allahabad High Court appears increasingly willing to ask those questions.

And in encounter cases, those questions may matter more than the encounter itself.