A covert campaign rarely announces itself. There is no declaration, no press briefing, no flag planted on a battlefield. What there is, instead, is a sequence of events that look unrelated until someone places them side by side and notices the rhythm. Three such events occurred across nine months in Pakistan. A car bomb detonated outside a Lahore neighborhood in June 2021. A wanted man fell to gunfire in January 2022. An aircraft hijacker living under a false name died inside a Karachi furniture shop in March 2022. None of the three carried a signature. Each was treated, when it happened, as a local crime or an isolated act of violence. Read in sequence, they form the opening chapter of the most consistent extraterritorial elimination campaign in modern South Asian history, and the moment a doctrine that had existed only as ambition crossed into practice.

This is the story of how the killing started. Not how it was conceived, because conception is invisible and leaves no record, but how it became real, observable, and repeatable. The argument running through this account is straightforward and contestable, and it deserves to be stated at the outset rather than smuggled in at the end. The shadow war did not begin with a single dramatic strike. It began with a car bomb that never reached the man it was apparently meant to kill, followed by two shootings that no one connected at the time, and by the middle of 2022 the campaign existed as a functioning instrument of statecraft even though no government had named it and no analyst had yet drawn the line through all three points. The June 2021 explosion near Hafiz Saeed’s Lahore residence functioned as a declaration of intent. The January 2022 killing of Saleem Rehmani functioned as a proof of concept. The March 2022 killing of Zahoor Mistry functioned as the confirmation that a pattern, not a coincidence, was now in motion. Understanding why those three events belong together, and why they mark a genuine break from everything that came before, is the work of this chapter in the longer chain that runs from the Kandahar tarmac in 1999 to the present.

The Preceding Link

Every event in this long chain inherits the conditions created by the event before it, and the covert campaign of 2021 and 2022 is no exception. To understand why the killing started when it did, the place to begin is the constitutional decision taken in New Delhi in August 2019, when the Indian Parliament revoked the special status that the territory of Jammu and Kashmir had held since 1954. That decision, examined in detail as the preceding political shift, did more than rearrange a domestic legal framework. It rewrote the strategic relationship between two nuclear-armed states and removed several of the assumptions that had governed how each side calculated risk.

For nearly three decades before 2019, the contest over Kashmir had operated within a set of unwritten rules. Pakistan-based militant organizations infiltrated fighters across the Line of Control and conducted attacks inside Indian territory. India responded with internal counter-insurgency operations, diplomatic protest, and occasionally with cross-border raids that both governments preferred to leave ambiguous. The militant leadership lived openly in Pakistani cities. Hafiz Saeed addressed rallies. Masood Azhar issued statements. The men who planned attacks against India did not hide, because hiding was unnecessary; the Pakistani state offered them sanctuary, and the Indian state, whatever its frustration, did not reach across the border to touch them. That arrangement had a name in the strategic literature, even if the public never used it. It was a stable equilibrium, ugly and lethal, but stable.

The revocation of Article 370 broke the equilibrium in a way that the men who drafted the legislation may not have fully intended. By integrating Jammu and Kashmir fully into the Indian Union, New Delhi signaled that it no longer regarded the territory’s status as negotiable, and therefore no longer regarded Pakistan as a party whose grievance required management. Pakistan responded by downgrading diplomatic relations, suspending bilateral trade, and appealing to the United Nations Security Council, where the appeal produced sympathy from China and very little else. The diplomatic track, which had always been thin, effectively closed. What remained was a relationship with no functioning channel for de-escalation and no shared framework for limiting the contest. When two states stop talking, the contest does not stop. It moves into spaces where talking was never required.

There is a second consequence of 2019 that matters even more directly for the events of 2021 and 2022. The revocation, combined with the lockdown and communications blackout that followed it inside the Kashmir Valley, hardened a conviction inside the Indian security establishment that had been building since the Pulwama attack and the Balakot airstrike earlier that same year. The conviction was this: that conventional military retaliation against Pakistan, however satisfying in the moment, carried escalation risks that capped its usefulness, and that the militant infrastructure responsible for attacks on India would never be dismantled by Pakistan voluntarily and could not be dismantled by Indian airpower without risking a wider war. If the infrastructure was to be degraded, it would have to be degraded at the level of the individual. The leadership, the planners, the financiers, the recruiters, the men whose names appeared on India’s most-wanted lists, would have to be reached where they lived. That is a covert mission by definition. It cannot be announced, because announcing it invites both retaliation and international censure. It can only be done quietly, deniably, and patiently.

So the period between August 2019 and June 2021 should be understood not as a lull but as a transition. Roughly twenty-two months passed between the revocation and the Lahore car bomb. During that interval, no major Indian operation against Pakistan-based targets became public. The absence of visible activity led some observers to conclude that India had returned to its older posture of restraint, the posture examined in the analysis of the campaign overview that frames the entire shadow war as a single coherent doctrine rather than a series of accidents. That conclusion was wrong, and the events of mid-2021 onward demonstrate why. Restraint, in the years after Balakot, did not mean inaction. It meant the substitution of one mode of action for another. The loud instrument was being set down. A quieter one was being picked up. The interval of apparent silence was the interval of preparation, and preparation for a campaign of targeted eliminations on foreign soil is itself a major undertaking. It requires intelligence collection, the identification and physical location of targets, the recruitment or cultivation of assets capable of operating inside Pakistani cities, the establishment of logistics and escape routes, and the construction of deniability robust enough to survive scrutiny. None of that happens overnight. The twenty-two months were not empty. They were full of work that, by design, produced no headlines.

Worth dwelling on, too, is the question of why a constitutional decision about Kashmir’s legal status would translate into a covert campaign at all. The connection is not obvious on its surface, and skeptics are right to ask for the mechanism rather than accepting a vague correlation. The mechanism runs through Pakistan’s own response. When Islamabad downgraded relations, expelled the Indian high commissioner, and suspended bilateral trade, it did not merely register protest. It removed the channels through which the two states had historically absorbed shocks. A diplomatic relationship, even a hostile one, functions as a pressure valve: it gives each side a way to signal limits, to communicate that a particular action will provoke a particular response, and to climb down from a confrontation without losing face. With that valve closed, the relationship lost its capacity to self-regulate. Both capitals were now operating with less information about the other’s intentions and fewer ways to communicate restraint. In such conditions, covert action becomes more attractive precisely because the alternatives have narrowed. There is no negotiating table at which to extract concessions, and the conventional military option remains capped by escalation risk. The covert track is what remains, and the 2019 revocation, by closing the other tracks, made the covert track the default rather than the exception.

A second strand of the inheritance concerns the militant organizations themselves and their reading of the post-2019 environment. The revocation, and the security clampdown that accompanied it inside the Kashmir Valley, was interpreted by Lashkar-e-Taiba, Jaish-e-Mohammed, and the smaller groups as a provocation that demanded a response. The emergence of new front organizations in the months that followed, designed to give old groups deniable new faces, signaled that the militant infrastructure intended to remain active and to adapt rather than to stand down. From the Indian security establishment’s vantage point, this meant the threat was not receding and would not recede on its own. If the organizations were regrouping under new names and the Pakistani state was sheltering that regrouping, then the case for reaching into the sanctuary and degrading the human core of those organizations grew stronger, not weaker. The post-2019 period therefore contained two parallel processes that fed each other: the militant networks adapting to remain lethal, and the Indian establishment concluding that only individualized, sustained pressure could counter that adaptation. The opening sequence of the shadow war is the point at which the second process produced its first visible results.

Such, then, is the inheritance that the shadow war’s opening events received from the link before them. A closed diplomatic channel. A security establishment convinced that the individual, not the airstrike, was the correct unit of action. And a preparation period long enough to build the apparatus that the campaign would require. When the first event arrived in June 2021, it did not arrive out of nowhere. It arrived as the first visible output of a process that had been running, unseen, since the constitutional decision two years earlier.

What Happened

The campaign’s opening is best reconstructed as three discrete events, examined in the order they occurred, because the order is the argument. Each event built on the one before it, and the meaning of the sequence is not contained in any single act but in the progression from the first to the third.

The Lahore car bomb of June 2021

At approximately eleven o’clock on the morning of June 23, 2021, a car packed with an estimated thirty kilograms of explosives detonated in the Johar Town locality of Lahore, Pakistan’s second-largest city. The blast killed three people and wounded more than twenty, with at least six of the injured listed in critical condition at the government-run Jinnah Hospital. Television footage from the scene showed smashed windows, blown-open doors, and structural damage to a row of houses near the epicenter. The Counter Terrorism Department of Punjab Police sealed the site within hours.

The detail that transformed a deadly bombing into a strategic event was the location. The damaged houses sat near the residence associated with Hafiz Muhammad Saeed, the co-founder of Lashkar-e-Taiba and the chief of its charitable front, Jamaat-ud-Dawa. Saeed is the man India and the United States hold responsible for masterminding the 2008 Mumbai attacks that killed more than one hundred sixty people. He is a United Nations designated terrorist, carries a ten million dollar United States bounty, and at the time of the blast was serving multiple sentences for terror financing in a Lahore jail. A spokesman for Jamaat-ud-Dawa told reporters that Saeed was in prison and therefore not in the residence that may have been the target. The bomb, in other words, struck near the home of one of the most consequential terrorists alive, on a morning when the man himself was not there to be struck.

Pakistani authorities moved quickly, both operationally and politically. Within two days, raids produced arrests, and police announced that those involved had been detained, including an Afghan national who had reportedly parked the explosives-laden vehicle at the site. On July 4, 2021, Pakistan’s National Security Adviser, Moeed Yusuf, held a news conference in Islamabad and accused India directly. He stated that forensic analysis and recovered electronic equipment had identified the mastermind and the handlers of the attack, that the mastermind belonged to India’s external intelligence service, the Research and Analysis Wing, and that the man was an Indian citizen based in India. Yusuf added that the bombing had coincided with what he described as thousands of coordinated cyberattacks on Pakistani investigative infrastructure, which he offered as further proof of state involvement. India rejected the accusation.

Several investigative details that surfaced in the weeks afterward deserve attention, because they bear on the question of how much preparation the operation required. The damaged residence sat in the Builders and Operatives Research Society area of Johar Town, a planned neighborhood, and the explosive vehicle was positioned near a police picket that had been stationed outside the high-profile residence precisely because of who lived there. Punjab’s provincial police chief observed that the picket itself had absorbed part of the blast and that the casualty toll, severe as it was, might have been higher without it. Reporting in the following months indicated that investigators detained and questioned a foreign national who had been moving frequently between Karachi, Lahore, and Dubai, and who was offloaded from a Karachi-bound flight for interrogation. Whatever the ultimate accuracy of Pakistan’s specific claims, the picture that emerges from the investigation is of an operation with a logistical footprint: a vehicle, roughly thirty kilograms of explosive material, local handlers, knowledge of a guarded address, and the ability to position the device near a police picket without being intercepted beforehand. That footprint is the evidentiary heart of the later argument that the campaign’s preparation predated the explosion.

For the purposes of this account, the question of who built and placed the bomb matters less than the question of what the bomb meant, and the two questions should not be collapsed into each other. The car bomb is the noisiest event in the entire opening sequence. It killed civilians. It produced rubble and televised footage. It generated a formal accusation from a national security adviser. In every respect it resembles the older mode of cross-border violence, the mode of spectacle and mass casualty, far more than it resembles the precise, quiet shootings that would follow. That apparent contradiction is exactly why the bomb belongs at the start of the story and not outside it. The detailed reconstruction of the Johar Town blast as the declaration of intent treats it as a foreshadowing event, a strike that reached toward the apex of Lashkar-e-Taiba’s leadership and missed, and whose significance lies less in what it destroyed than in what it signaled about where the campaign intended to go.

One further point about the bomb deserves to be made before moving on, because it shapes how the rest of the sequence reads. If a covert campaign were designed from the outset around the model of quiet, deniable shootings, the car bomb would be an odd opening move, since it is the opposite of quiet and deniable. There are two plausible readings of that mismatch, and both are instructive. The first reading is that the bomb was a relatively early operation, conducted before the campaign had settled on the shooting model, and that its noise reflects an instrument still being calibrated. The second reading is that the bomb and the shootings were never meant to be the same kind of act: the bomb a one-time attempt to reach an apex target by the only means that could plausibly reach a man inside a guarded compound, the shootings a sustainable method for the many targets who live more exposed lives. Both readings support the same conclusion, which is that the opening sequence is not three identical operations but a progression in which an actor learns, adapts, and converges on the method it will keep. The bomb is the part of that progression that the actor did not repeat.

The killing of Saleem Rehmani in January 2022

Roughly seven months after the Lahore blast, in January 2022, a man named Saleem Rehmani was shot dead in Pakistan by unidentified gunmen. Rehmani appeared on India’s list of designated and wanted individuals. Open-source reporting on him is thin, which is itself a significant fact and one this account will return to, but the essential outline is clear enough. He was a man India sought, he was living in Pakistan, and he was killed by assailants whose identity was never established and whose act no organization claimed.

The thinness of the Rehmani record stands in deliberate contrast to the volume of coverage the Lahore bomb received. The car bomb was a public event. It could not be hidden, because it announced itself with sound and rubble in the middle of a major city. The Rehmani shooting was the opposite. It produced no televised footage of structural damage, no national security adviser at a podium, no immediate forensic narrative. A wanted man died, and the event passed with comparatively little notice. It entered the public record mostly in retrospect, when later reporting, including the assessments compiled in international investigations of the broader pattern, listed Rehmani among the earliest cases in a sequence of similar killings.

That low profile is precisely what makes Rehmani important. If the Lahore bomb was an attempt at spectacle, the Rehmani killing was an experiment in invisibility. It tested whether a wanted man could be located inside Pakistan, reached, and eliminated without the operation generating attribution, evidence, or international consequence. The fuller treatment of Rehmani as the proof of concept makes the case that his death matters in the chain not because of who he was, since he was a mid-tier figure rather than a household name, but because of what his killing demonstrated. It demonstrated that the method worked. A man on India’s list could be removed, and the removal could look, to the casual observer and even to the attentive one, like an ordinary act of Pakistani street violence. The campaign, if a campaign existed, had found a form that left almost no surface for accusation to grip.

There is a useful way to understand why a campaign would choose to follow a loud event with a deliberately quiet one. A car bomb that misses its target has, in a sense, spent capital without buying the intended result. It has alerted the adversary, hardened the protection around apex figures, and invited diplomatic blowback, all without removing anyone of consequence. An actor that learns from such an outcome would reasonably conclude that the next move should minimize every one of those costs. The Rehmani operation does exactly that. It produces no warning that ripples across the militant leadership, because the target is obscure and the method is silent. It generates no diplomatic crisis, because there is nothing dramatic enough to crystallize one. And it removes an actual name from India’s list, which the bomb did not. Read this way, the progression from the Lahore bomb to the Rehmani killing is not random variation. It is the visible trace of an actor correcting course, retaining the ambition the bomb declared while discarding the noise that made the bomb costly. The campaign did not merely continue after June 2021. It adjusted.

There is one further reason the Rehmani case rewards attention despite its thin record. Mid-tier figures, the recruiters and launching coordinators and logistics men who occupy the layer beneath the famous leadership, are in operational terms often more valuable to degrade than the figureheads. A movement can survive the imprisonment of a celebrity ideologue, because ideology is diffuse and replaceable. It survives less easily the steady removal of the practical men who actually move fighters, weapons, and money across borders, because that expertise is concrete, specific, and slow to rebuild. If the campaign began its confirmed killings at this layer, the choice is consistent with an actor that understood where a militant network is genuinely vulnerable. The obscurity that makes Rehmani a minor figure in the headlines is not the same as insignificance in the architecture of a terror network, and the campaign’s willingness to spend its second operation on such a target suggests a clear-eyed reading of that distinction.

The killing of Zahoor Mistry in March 2022

The third event arrived on March 1, 2022, in Karachi, and unlike the first two it attached itself to a name that the Indian public recognized instantly. Zahoor Mistry, also recorded in some accounts as Mistry Zahoor Ibrahim and known by the alias Zahid Akhund, was shot twice at point-blank range in the head by two assailants who arrived and departed on a motorcycle. The killing took place in the Akhtar Colony area of Karachi, and reporting indicated that he was struck inside the premises of a furniture business he ran under his assumed identity. Closed-circuit footage reportedly showed two helmeted, masked men conducting reconnaissance of the area before the shooting. The news of his death reached India in the second week of March, several days after the act itself.

Mistry was not a mid-tier figure. He was one of the five men who hijacked Indian Airlines Flight IC-814 in December 1999, the hijacking that forced India to release three imprisoned terrorists, among them Masood Azhar, in exchange for nearly two hundred hostages. The full account of that hijacking, traced in the IC-814 hijacking complete guide, establishes Mistry as one of the most reviled individuals in the modern Indian memory of terrorism. During the hijacking he used the code name doctor, and he is identified as the man who stabbed a young passenger, Rupin Katyal, who was returning from his honeymoon, leaving him to bleed to death aboard the aircraft. After the crisis ended, Mistry vanished into Pakistan, took the name Zahid Akhund, built a furniture company, and lived for more than two decades in Karachi under that false identity, reportedly with the protection of the Pakistani intelligence apparatus.

To grasp the weight of Mistry’s killing, the reader has to hold the IC-814 hijacking in view, because Mistry’s entire significance derives from it. The aircraft, carrying nearly two hundred people, was seized shortly after takeoff from Kathmandu in December 1999 and flown through Amritsar, Lahore, and Dubai before coming to rest at Kandahar, in an Afghanistan then controlled by the Taliban. Over the days that followed, India negotiated the hostages’ freedom in exchange for the release of three men from Indian custody, a transaction whose consequences this entire chain exists to trace. Mistry was not the planner of that operation, but he was its sharpest edge: the man on the aircraft, the man with the knife, the man whose act of murder gave the hijacking its most enduring image of cruelty. For twenty-three years he lived beyond the reach of Indian justice, in plain sight under a new name, running a business and raising the ordinary scaffolding of a settled life. The gap between the enormity of what he had done and the comfort of how he was living was, for the Indian public, an open wound. His killing did not heal that wound, but it closed a particular kind of impunity that had seemed permanent.

His funeral confirmed how connected he remained. Reporting indicated that senior figures of Jaish-e-Mohammed attended the prayers, including the organization’s operational leadership, and that the network founded by the man Mistry had helped free in 1999 turned out in strength to bury him. Among those reported to have attended was the operational chief of the organization, a brother of the Jaish-e-Mohammed founder, a detail that links Mistry’s death directly back to the IC-814 transaction that created the organization in the first place. With his death, only two of the five IC-814 hijackers were believed to remain alive, both of them close relatives of the Jaish-e-Mohammed founder. Pakistan again attributed the killing to Indian intelligence, alleging that New Delhi had hired assassins for the act.

The reconstruction of Mistry’s life and death as the pattern confirmation draws out why this third event closed the opening sequence. Mistry’s killing combined the two qualities that the first two events had displayed separately. It had the precision and quietness of the Rehmani shooting: two men, a motorcycle, head shots at close range, masks, no claim, no attribution that could be proven. And it had a target whose significance approached, though it did not equal, the significance the Lahore bomb had reached for. Mistry was a name. His death was reported across Indian media. The method that had been tested quietly on a mid-tier figure in January was now applied to a man the Indian public had spent twenty-three years hating, and it worked exactly as it had worked before. The motorcycle, the masks, the close-range shooting, the absence of any claim: this was now visibly a method, not an accident. By the morning of March 2, 2022, anyone willing to lay the three events on a single timeline could see the shape of something.

The findable artifact: the initiation phase as declaration, proof, and confirmation

Taken together, the three events of June 2021, January 2022, and March 2022 are best understood not as a list but as a structured progression, and that progression is the distinctive analytical artifact this account contributes to the longer chain. Each event performed a function, and the functions are sequential. Naming the functions precisely is what turns a chronology into an argument.

Consider the Lahore car bomb first. It was the declaration of intent. It was loud, it was aimed at the apex of Lashkar-e-Taiba’s leadership, and it failed in its apparent immediate purpose because the man it reached toward was in jail. A declaration does not need to succeed in order to declare. Its content is the ambition it reveals, and the ambition the Johar Town blast revealed was the willingness to strike at the very top of the militant hierarchy inside a major Pakistani city. What the bomb declared, the campaign would spend the next several years trying to deliver more precisely. The declaration also carried a message to the host state, namely that the sanctuary it provided could be entered and that even its most valuable guest was not categorically beyond reach.

Rehmani’s killing was the proof of concept. It took the ambition declared in Lahore and tested whether it could be executed quietly. It answered a narrow operational question: can a wanted man inside Pakistan be located and eliminated without producing attributable evidence or international consequence. The answer was yes. A proof of concept is rarely glamorous and rarely famous, and Rehmani’s near-invisibility in the public record is consistent with that role. A proof of concept is meant to be observed by the people running the experiment, not by the public. Its success or failure is measured not by headlines but by whether the method can now be trusted.

Mistry’s killing, finally, was the confirmation of pattern. It applied the validated method to a target of genuine public significance and produced the same result by the same means. Confirmation is the moment at which a method stops being an experiment and becomes a doctrine, because it has now worked twice, against different kinds of targets, by an identical procedure. After Mistry, the campaign no longer needed to ask whether the method worked. It had been confirmed. The remaining questions were about scale, tempo, and target selection, and those questions belong to the years that followed. A pattern, once confirmed, becomes a template, and a template is what allows a campaign to grow.

Declaration, proof, confirmation. That is the architecture of the shadow war’s opening, and it is an architecture that the loud-then-quiet-then-quiet-with-a-name sequence makes visible. A campaign that the rest of this long chain documents did not spring into existence fully formed. It was assembled, in public view, across nine months, by an actor patient enough to let each event do only its own job. The value of stating the architecture explicitly is that it converts what looked, at the time, like three unrelated misfortunes into what it almost certainly was: the deliberate, staged commissioning of a new instrument of statecraft.

Why It Happened

A reconstruction of what happened is incomplete without an account of why it happened in this form, at this time, and by this method. Three causal questions deserve direct answers: why a covert campaign rather than a conventional one, why the years 2021 and 2022 specifically, and why this particular operational signature of motorcycles, masks, and unclaimed shootings.

The first question, why covert rather than conventional, has its answer in the recent military history that this chain has already traced. The Balakot airstrike of February 2019 was the most ambitious conventional Indian action against Pakistani territory in nearly five decades. It crossed the international boundary, not merely the Line of Control, and it struck what India described as a Jaish-e-Mohammed training facility. The strike was followed by an aerial engagement in which an Indian pilot was captured and then returned. Balakot was, in the Indian domestic narrative, a demonstration of resolve. But it also demonstrated a limit. A conventional strike against Pakistan produces an immediate Pakistani conventional response, and the exchange escalates along a ladder whose upper rungs both states have powerful reasons to fear. Balakot showed that India could strike conventionally and survive the consequences, but it also showed that conventional striking is a blunt instrument that cannot be used often, cannot be calibrated finely, and carries the permanent risk of an escalation that neither capital can fully control. If the goal is to degrade militant infrastructure steadily over years, the conventional strike is the wrong tool. It is too loud, too dangerous, and too infrequent. A campaign that wants to work continuously must work quietly.

There is a deeper logic underneath the escalation argument. The militant infrastructure that India wished to degrade was not, fundamentally, a collection of buildings. Buildings can be rebuilt within months, and training camps are notoriously easy to reconstitute. The infrastructure was a collection of people: the planners who designed attacks, the recruiters who filled the ranks, the financiers who moved the money, the hijackers and masterminds whose survival and freedom themselves served as recruitment advertisements. The analysis of Pakistan’s terror safe haven network makes the point that the sanctuary Pakistan offered was valuable to the militant organizations precisely because it kept irreplaceable human expertise alive and functioning. A bomb dropped on an empty camp degrades nothing of lasting value. The removal of a specific planner, a specific hijacker, a specific financier, degrades something that cannot be quickly replaced. If the unit of value is the individual, the unit of action must also be the individual, and reaching individuals where they live is a covert mission. The choice of method followed from a correct identification of the target.

The second question, why 2021 and 2022 specifically, returns the analysis to the preparation interval discussed earlier. The capacity to conduct targeted eliminations inside Pakistani cities is not a switch that can be flipped. It is an apparatus that must be built, and building it takes time. Intelligence on the precise current location of a man living under a false identity in Karachi does not exist by default; it must be collected, verified, and kept current, because targets move. Assets capable of conducting an operation inside a hostile city must be identified, recruited, trained, positioned, and given workable escape routes. Deniability must be engineered into every layer, because an operation that can be traced back to its sponsor is an operation that produces exactly the international consequence the covert method exists to avoid. The roughly twenty-two months between the August 2019 revocation and the June 2021 bomb is consistent with the time such construction would require. The campaign became visible in 2021 because that is approximately when the apparatus reached the threshold of operational readiness. It did not become visible earlier because earlier the apparatus did not yet exist. The dating of the opening sequence is therefore not arbitrary. It reflects the lead time of building a covert capability from a standing start.

The third question concerns the operational signature, and here the answer is partly about effectiveness and partly about deniability. The motorcycle is the ideal urban assassination platform. It moves quickly through congested streets, it does not require the assailants to expose themselves the way a car forces a driver to, it can be abandoned without forensic cost, and it is so common on South Asian roads that it draws no attention before the act. The mask and helmet defeat closed-circuit identification, which matters enormously in cities now saturated with cameras. The close-range head shot guarantees the outcome and minimizes the number of rounds, noise, and time spent at the scene. The absence of any claim of responsibility is the most strategically important feature of the signature, and it is a feature, not an omission. A claimed killing invites retaliation and international censure. An unclaimed killing leaves the target’s organization and the host state with a problem they cannot easily externalize, because they cannot prove who did it, and proving who did it is the precondition for any formal response. The study of the unknown gunmen pattern treats this signature as the deliberate core of the doctrine: every element of the method is chosen to maximize the gap between what observers can reasonably infer and what they can actually prove.

There is one more element of the why that the opening sequence reveals, and it concerns the choice of targets in this specific order.

Before turning to that ordering, one further dimension of the covert choice deserves examination, because it concerns cost and sustainability rather than escalation risk. A conventional military campaign is extraordinarily expensive, not only in money but in the political capital that must be spent each time the instrument is used. Every airstrike requires a domestic justification, invites international scrutiny, and consumes diplomatic energy that could be spent elsewhere. It is, by its nature, an episodic instrument: a state can mount a Balakot, but it cannot mount a Balakot every quarter without exhausting itself. A covert campaign of targeted eliminations has the opposite cost structure. Each individual operation is comparatively cheap, and because it is deniable, it consumes almost no political capital, since the sponsoring state never has to defend in public an act it never admits. Over a span of years, this difference compounds. The covert method allows a state to apply pressure continuously, week after week, at a cost it can sustain indefinitely, whereas the conventional method allows only occasional, expensive spasms of pressure separated by long intervals of inaction. If the strategic goal is the steady degradation of an adversary’s militant infrastructure rather than a single dramatic punishment, the covert method is not merely safer. It is the only method that the budget of political capital can actually afford to run for the length of time the goal requires.

A related consideration concerns the management of domestic expectations, and it cuts in a direction that is easy to miss. A government that retaliates conventionally and visibly raises the public’s expectations each time it acts, because a visible strike invites the question of what comes next, and a public that has been shown a spectacular response will demand another after the next attack. Visible retaliation, in other words, traps a government on an escalating ladder of its own making. A covert campaign escapes that trap. Because the campaign is never acknowledged, the government incurs no public obligation to escalate, and it retains the freedom to calibrate the campaign’s tempo to operational reality rather than to the public mood. This freedom is strategically valuable, and it is part of why a security establishment that had just demonstrated, at Balakot, that it could retaliate visibly might nonetheless choose, for the longer campaign, an instrument that the public would never see. The covert method is not only cheaper and safer. It also leaves the government’s hands freer.

Returning now to target selection, the campaign did not begin by going after the most prominent living terrorist it could reach. Hafiz Saeed himself was alive, and although imprisoned, the Lahore bomb showed that the campaign was willing to reach in his direction. But the bomb missed, and the campaign did not immediately try again to reach the apex. Instead, the next two confirmed killings descended the ladder of prominence in a way, going to a mid-tier figure first and then to a hijacker whose fame rested on a single notorious episode rather than on current organizational command. This is consistent with an actor learning its own method before betting it on the hardest targets. You do not test a new instrument on the most heavily defended objective. You test it where failure is survivable, you confirm that it works, and you reserve the apex for later, when the method is proven and the apparatus is mature. The order of the opening targets is the order of an organization climbing a learning curve, and that, more than anything, is what marks the sequence as the deliberate beginning of a campaign rather than a string of unrelated deaths.

The named disagreement: did the campaign begin with the bomb, or before it?

An honest account of why the shadow war began in 2021 must engage a genuine analytical dispute, because the dating of the beginning is not settled. One position holds that the June 2021 car bomb is the true start, the first operation of the campaign and therefore its origin point. A competing position holds that the campaign began considerably earlier, that its real origin lies in the intelligence preparation that the bomb itself required, and that to date the campaign from its first visible explosion is to confuse the moment a process became observable with the moment it began.

The second position has the stronger logic, and the reasoning is worth following carefully. A thirty-kilogram vehicle-borne explosive device placed near a specific high-value residence in a specific neighborhood of Lahore is not an improvised act. It presupposes a great deal. It presupposes that someone knew which residence to target and where it sat. It presupposes a local network capable of acquiring a vehicle, obtaining and assembling explosives, and positioning the device at the chosen spot. It presupposes that this network could operate inside Lahore, one of the most heavily policed cities in Pakistan, without being detected before the act. None of that infrastructure assembles itself in the days before a bombing. It is the product of months of patient collection, recruitment, and logistics. If the campaign is defined as the operational apparatus rather than as its first detonation, then the campaign was clearly running well before June 23, 2021. The bomb was not the birth of the campaign. It was the campaign’s first visible output.

This is where the testimony of analysts who study the Indian security establishment becomes useful, and two in particular have addressed the question directly. The defense journalist Saikat Datta has examined when the intelligence community first recognized that a coherent pattern was forming, and his work points to the gap between operational reality and analytical recognition, the period during which a campaign is functioning but has not yet been named even by the professionals whose job is to name such things. The counter-terrorism analyst Praveen Swami has written extensively on the campaign’s initiation phase, and his analysis emphasizes that India’s reorientation toward covert action against Pakistan-based targets was a considered strategic shift with roots that predate any single incident, rather than a reaction improvised after a particular attack. Taken together, these assessments support dating the campaign’s true beginning to the preparation period and treating the bomb as the first event the public could see.

The dispute over dating is not merely a scholar’s quarrel about calendars, because the answer one gives shapes how the entire campaign is understood. If the campaign began with the bomb, then it is most naturally read as a reactive enterprise, something that started in response to a provocation and built from there. If instead the campaign began in the quiet preparation that the bomb required, then it is better read as a deliberate strategic program, conceived as a program from the outset, with the bomb as a scheduled early output rather than a starting gun. The second reading carries heavier implications. It means the covert turn was a settled element of policy, planned and resourced in advance, rather than an angry improvisation. It means the apparatus was designed to run for years, because no state builds a months-long intelligence and logistics effort to produce a single result. And it means that the events the public would later watch, the killings of 2022 and the faster killings of the year after, were never separate decisions but the unfolding of a plan whose first chapter was written before any explosion. Dating the campaign to its preparation, in other words, is what licenses the reader to treat the whole later sequence as a campaign at all. That is why the disagreement matters, and why this account, while presenting both positions fairly, leans toward the earlier date.

The two positions can be reconciled by being precise about what is being dated. If the question is when did the campaign first become observable to the outside world, the answer is June 23, 2021, the day of the Lahore bomb. If the question is when did the campaign begin as an enterprise, the answer is earlier, somewhere in the long interval after the 2019 revocation, when the decision was taken and the apparatus began to be built. This account uses the bomb as the opening event because a chain of events must be told through events, and the bomb is the first event there is. But the reader should hold the larger truth alongside the narrative convenience. The shadow war’s opening sequence is the visible portion of an iceberg whose larger mass, the preparation, the recruitment, the collection, lies beneath the waterline and produced no events for a chain to record.

The complication that honesty requires

There is a limit to how firmly any of this can be asserted, and the limit must be stated plainly rather than buried. The claim that the June 2021 bomb, the January 2022 shooting, and the March 2022 shooting belong to a single campaign is an analytical inference. It is not a confirmed fact, and it cannot be, given the nature of covert operations and the absence of any acknowledgment by any government.

The inference rests on pattern. Three events, against targets that India sought, inside Pakistan, within nine months, with no organization claiming any of them, and with the second and third sharing an operational signature, form a shape that is difficult to attribute to coincidence. But pattern is not proof. It remains possible, in principle, that the Lahore bomb and the two shootings had different sponsors, or that one or more of them resulted from internal Pakistani militant rivalries, score-settling within or between organizations, or criminal disputes unconnected to any state campaign. Pakistan’s own attribution to Indian intelligence is an accusation from an interested party, not independent confirmation, and an interested party’s accusation cannot be treated as evidence merely because it happens to align with the inference. India’s denials are equally the statements of an interested party.

Honesty therefore requires holding the argument of this account at the correct strength. It is a strong inference, well supported by the convergence of target selection, geography, timing, signature, and the absence of claims, and it is the inference that best explains the full set of facts. It is not a certainty. The reader who wants certainty about covert operations will wait a long time, because covert operations are designed precisely to deny it. What can be said with confidence is narrower and still substantial: that by the middle of 2022, India’s most-wanted individuals living in Pakistan were dying by violence at a rate and in a manner that had no precedent in the previous decade, and that whatever the explanation, the security environment for those individuals had changed fundamentally. The campaign interpretation is the best available reading of that change. It should be carried forward as a strong working thesis, not as a proven fact.

The Immediate Consequences

The opening sequence produced consequences within Pakistan, within India, and within the militant organizations themselves over the weeks and months that followed, and these immediate effects matter because they shaped the conditions under which the campaign would expand.

Inside Pakistan, the most visible immediate consequence was the public accusation. Moeed Yusuf’s July 2021 press conference was not a minor diplomatic note; it was a formal, on-the-record charge by a national security adviser that a neighboring state had sponsored a lethal bombing on Pakistani soil. Pakistan repeated and elaborated the accusation in December 2022, presenting what it described as evidence of Indian involvement in the Johar Town blast. This response is revealing in two directions at once. It shows that the Pakistani state took the bombing seriously and was prepared to spend diplomatic capital on it. It also shows the structural weakness of the position the campaign had placed Pakistan in. An accusation, however confidently delivered, is not the same as a consequence. Pakistan could name India, but naming produced no penalty, no international action, no reversal. The accusation became, in effect, an admission that something was being done to Pakistan that Pakistan could not stop and could not make the world care about. For an audience watching closely, the July 2021 press conference was less a demonstration of Pakistani strength than a demonstration of Pakistani exposure.

The killings of Rehmani and Mistry produced a quieter but more consequential set of effects inside the militant organizations. The death of a hijacker who had lived safely in Karachi for more than twenty years under an assumed identity, protected, by the standard account, by the intelligence apparatus, carried a message that every comparable figure in Pakistan would have received immediately. The message was that the false identity was no longer sufficient. The furniture business, the new name, the years of quiet residence, the protection: all of it had failed to keep Zahoor Mistry alive. The men who had built similar lives, the planners and operatives and financiers living under cover in Pakistani cities, now had concrete reason to believe that their cover could be penetrated and that penetration meant death. The organizational guides that trace the internal histories of Lashkar-e-Taiba and Jaish-e-Mohammed describe a measurable tightening of security behavior among senior figures in the period after these early killings, a shift toward reduced public movement, more frequent relocation, smaller and more trusted circles, and greater caution about predictable routines. That behavioral change is itself an immediate consequence of the opening sequence, and it is a consequence with strategic weight. An organization whose leaders must spend their attention on personal survival is an organization spending less attention on planning attacks.

There was also an immediate psychological consequence for the relationship between the militant organizations and their state sponsor. Sanctuary is a transaction. The militant organizations provided the Pakistani security establishment with deniable instruments of pressure against India, and in return the state provided protection, including the protection of physical safety. When that protection visibly failed, when a hijacker under intelligence cover was shot in a Karachi furniture shop, the value of the transaction came into question from the militants’ side. The state could not deliver the safety it had implicitly promised. This did not break the relationship, which rested on too many other foundations, but it introduced a strain into it, a doubt about reliability, that had not been there before. The full account of Hafiz Saeed’s protected status shows how completely the apex of the militant leadership had come to depend on state protection, and the Lahore bomb, by reaching toward Saeed himself, demonstrated that even the apex was not beyond the campaign’s ambition. The opening sequence, in its immediate aftermath, began to erode the confidence on which the entire sanctuary arrangement depended.

A further immediate consequence registered inside the Pakistani security apparatus itself, and it took the form of a counter-intelligence problem that the apparatus could not easily solve. A killing carried out against a protected individual is also, implicitly, an indictment of the protection. When Mistry was shot, the question that the protecting agency had to confront was not only who killed him but how the killers had found him, because a man living for two decades under a false identity is found only if the secrecy that concealed him has been breached somewhere. That breach could lie in a compromised official, a turned associate, a careless pattern of movement, or a penetration of the very networks that were supposed to keep him hidden. Each of those possibilities is alarming to a security establishment, and the establishment could not know which one applied without an investigation that would itself consume resources and sow internal suspicion. The opening sequence therefore imposed a cost on Pakistan that went beyond the dead men and beyond the public humiliation of the accusations. It forced the protecting apparatus to spend its attention hunting for leaks inside itself, and an apparatus turned inward in search of traitors is an apparatus distracted from its outward mission. This is a quieter consequence than the press conference, and it left no documentary trace that an outside analyst can cite with confidence, but it follows logically from the nature of the killings, and it is part of the pressure the opening sequence began to apply.

Inside India, the immediate consequences were mostly characterized by silence, and the silence was deliberate. The Indian government did not claim the killings. It did not celebrate them in any official capacity. It maintained, then as later, that it had no role in extraterritorial assassinations. This official silence is itself an immediate consequence worth examining, because it is the operational requirement of the covert method made visible. The campaign cannot work if its sponsor acknowledges it, because acknowledgment converts a deniable operation into an act of state for which the state can be held to account. The silence was not a sign that nothing had happened. It was a sign that the method was being respected. There was, alongside the official silence, an unofficial current of public satisfaction inside India, particularly in the case of Mistry, whose role in the IC-814 hijacking and in the killing of Rupin Katyal had made him an object of national revulsion for more than two decades. That public mood, the quiet sense that a long-delayed account was being settled, was an immediate consequence of a different kind, one that operated on the level of national feeling rather than policy, and it would matter later as the campaign expanded, because a covert campaign that enjoys tacit domestic approval is a campaign that its government has little political reason to halt.

Finally, the opening sequence produced an immediate consequence in the realm of analysis and reporting, although this one matured more slowly. In the weeks after Mistry’s death, the three events still sat mostly separate in public understanding. The bomb was remembered as a bomb. Rehmani was barely remembered at all. Mistry was reported as the overdue death of a notorious hijacker. The line connecting them had not yet been widely drawn. But the raw material for drawing it now existed, and a small number of journalists and analysts had begun to notice the convergence. The recognition that these were not isolated events but the visible portion of a pattern would take many more months and several more killings to become a settled view, and the timeline of every targeted killing in Pakistan shows how the cases accumulated before the pattern crystallized in the public mind. But the accumulation started here, with these three events, and the immediate consequence of the opening sequence was to lay down the first three points through which a line would eventually be drawn.

The Long-Term Chain

The opening sequence of 2021 and 2022 mattered far beyond its immediate aftermath, because it established conditions, methods, and expectations that shaped everything the campaign did afterward. To see those long-term effects clearly, it helps to ask what became possible, and what became likely, because of the way the campaign began.

A first long-term effect is methodological. The opening sequence did not merely conduct three operations; it validated a template. The motorcycle, the masks, the close-range shooting, the absence of a claim, the selection of a wanted individual living in a Pakistani city as the target: this combination, tested across Rehmani and Mistry, became the standard operating procedure of the campaign. The study of the modus operandi of covert eliminations describes how remarkably consistent the method remained across the years that followed, and that consistency is itself traceable to the opening sequence. A method that has been confirmed does not need to be reinvented. It needs to be repeated, refined, and scaled. The campaign of later years is, at the level of technique, the campaign of 2022 applied again and again. The opening sequence’s most durable contribution was to settle the question of how the killing would be done, so that the campaign’s subsequent energy could be spent on the questions of whom and when.

This methodological inheritance deserves a closer look, because a validated template is more than a convenience. It is a form of stored knowledge that lowers the cost and the risk of every operation that follows. When an organization conducts its first operation, it pays a steep price in uncertainty. It does not know whether the assets it has recruited will perform under pressure, whether the escape routes it has planned will hold, whether the local network supporting the operation has been penetrated, whether the chosen weapon and approach will produce a clean outcome. Each of those uncertainties carries a probability of failure, and failure in a covert operation is catastrophic, because a botched killing produces a captured asset, a surviving witness, or a forensic trail, any of which can collapse the deniability the whole method depends upon. The opening sequence retired those uncertainties one by one. By the time Mistry was killed in Karachi, the apparatus had a body of confirmed practice behind it. The later campaign did not have to relearn what the opening sequence had already taught, and that is why it could move faster while taking, paradoxically, less risk per operation. A template, once proven, converts a dangerous improvisation into a routine, and routine is what allows a campaign to run for years.

The second long-term effect is the establishment of deniability as a proven shield rather than a hopeful theory. Before the opening sequence, the idea that India could conduct eliminations inside Pakistan without bearing international consequence was a hypothesis. After the opening sequence, it was demonstrated. Pakistan had accused India directly, publicly, and repeatedly, and the accusations had produced no penalty. No international body acted. No major power treated the killings as a crisis. The world’s response to Pakistan’s charge was, in practical terms, indifference. This was a discovery of enormous strategic value to the campaign, because it meant that the central risk of covert action, the risk of exposure leading to consequence, had been tested and found to be low. The campaign could expand because the opening sequence had proven that expansion was safe. Every killing that followed inherited the deniability that these first events had validated.

Tempo and confidence form a third long-term effect. An organization conducting its first operations proceeds cautiously, because it does not yet know whether its method works, whether its assets are reliable, whether its deniability holds. The opening sequence answered all three questions affirmatively. The method worked. The deniability held. The campaign had, by the middle of 2022, every operational reason to accelerate, and accelerate it did. The progression from the careful, learning-curve pace of the opening sequence to the far higher operational tempo of the following year is documented in the account of the next phase, the year in which the killings came so frequently, and against targets so significant, that the existence of a campaign became impossible for any honest observer to deny. That acceleration was not a new decision. It was the natural consequence of an opening sequence that had removed every reason for caution.

Confidence, in this context, is not a vague psychological notion but an operational asset with measurable effects. A campaign that is confident of its method can plan further ahead, commit assets earlier, and tolerate a wider range of targets, because it is no longer hedging against the possibility that the core technique will fail. The opening sequence converted the campaign from a tentative experiment, in which each operation was also a test of the experiment itself, into a going concern that could treat each operation simply as a task to be executed. That shift in posture is one of the least visible but most important things the years 2021 and 2022 accomplished. It is the difference between an organization that is still asking whether it can do something and an organization that has moved on to deciding how much of it to do.

A fourth long-term effect operates on the militant organizations and the sanctuary system itself. The opening sequence began a process of degradation that compounded over time. Each killing removed a specific individual whose expertise the organization had to replace, and replacement is slow and imperfect. More importantly, each killing reinforced the lesson that sanctuary in Pakistan was no longer safe, which forced the surviving leadership into the defensive crouch described earlier: less movement, more relocation, smaller circles, constant attention to personal security. An organization in that crouch is degraded even when no one in it has been killed recently, because its capacity for ambitious, complex, externally directed operations declines as its internal attention shifts to survival. The opening sequence started that compounding process. The detailed picture of India’s most-wanted figures still in Pakistan shows a leadership cohort that, year by year, grew smaller, more cautious, and harder pressed, and the origin of that trajectory is the demonstration, in 2021 and 2022, that the cohort could be reached.

It is worth dwelling on why degradation of this kind compounds rather than simply accumulating. A campaign that removes one planner has removed one planner; that is accumulation. But a campaign that removes one planner and, in doing so, forces every remaining planner to halve his movements, distrust his couriers, and abandon his predictable routines has done something larger, because the surviving planners are now less effective than they were before, even though they are still alive. The cost of the first killing is therefore paid twice: once in the lost individual and once in the diminished output of everyone who remains. As the killings continue, this second cost grows, because each new death deepens the fear and tightens the crouch. An organization can absorb the loss of individuals for a long time, replacing them imperfectly but persistently. It cannot as easily absorb the steady erosion of its collective tempo, its institutional memory, and its willingness to take the operational risks that ambitious external attacks require. The opening sequence, by proving that the cohort was reachable, set this compounding erosion in motion, and the erosion did more lasting damage than the body count alone would suggest.

There is a fifth long-term effect, more abstract but no less real, and it concerns the strategic relationship between the two states. The opening sequence inaugurated a mode of contest that had not previously existed in the India-Pakistan relationship at this scale and consistency. For decades the violence had flowed predominantly in one direction, from Pakistan-based organizations into India, with India responding internally or, rarely, with conventional strikes. The shadow war introduced a sustained, deniable, individually targeted counter-flow. This did not equalize the relationship, and it did not end the attacks on India, as later events in the chain make painfully clear. But it changed the structure of the contest. Pakistan now faced a cost on its own soil, imposed continuously, that it could neither prevent nor avenge nor even prove. The opening sequence is where that new structure began, and the rest of the chain unfolds within it. The founder whose 1999 release from Indian custody the IC-814 hijacking secured, and who went on to build a terror enterprise around himself, would watch the organization he created come under exactly the kind of patient, individualized pressure that the opening sequence first demonstrated was possible. The structural change, in other words, reached the very figures who had once seemed permanently beyond the reach of any Indian response, and it reached them because the opening sequence had proven that reach was achievable.

What became possible because of 2021 and 2022, then, is a great deal. A validated method. A proven shield of deniability. A justified acceleration. A compounding degradation of the militant leadership. And a restructured contest in which Pakistan, for the first time, bore a continuous and unanswerable cost. None of these was fully visible in March 2022. All of them were latent in the three events that had just occurred. The opening sequence was small in its body count and modest in its immediate noise, but it was large in what it set in motion, and the chain that follows is the working out of its consequences.

One further observation belongs here, because it concerns how the opening sequence should be weighed against the louder events that came later. There is a natural tendency, in any chain of events, to assign the greatest importance to the moments of greatest drama, and by that measure the opening sequence would rank low, since two quiet shootings and a bomb that missed its target cannot compete for attention with a year of rapid, high-profile eliminations. But importance and drama are not the same thing. The opening sequence is where the decisive uncertainties were resolved, and resolving an uncertainty is the kind of achievement that leaves little visible trace while changing everything that follows. A campaign that has proven its method, validated its deniability, and demonstrated its reach has done the hard, irreversible work, even if it has done so without spectacle. The years that came after were louder, but they were also, in a sense, easier, because they operated inside a set of possibilities that the opening sequence had already established. To understand the shadow war as a whole, the reader must resist the pull of the dramatic and recognize that its foundational chapter was also its quietest, and that the quiet was not a sign of small consequence but a feature of how foundations are laid.

The Next Link

Every link in this chain ends by naming the link that follows, because the argument of the whole series is that these events form a connected sequence rather than a scattered set. The opening sequence of 2021 and 2022 leads directly into the year that turned a quiet experiment into an undeniable campaign.

If 2021 and 2022 were the years of declaration, proof, and confirmation, the year that followed was the year of escalation. The careful, learning-curve pace of the opening sequence, two confirmed killings spread across the months after a foreshadowing bomb, gave way to something far faster and far harder to dismiss. Across the following year, a succession of wanted men died in Pakistan by the now-familiar method, and they died not at intervals of months but in a steady, accelerating rhythm, against targets that climbed back up the ladder of organizational seniority that the opening sequence had cautiously descended. The number of killings in a single year reached a level that no coincidence could plausibly explain, and the geographic spread widened from the Punjab heartland into Pakistan-administered Kashmir and beyond. The account of the year of eliminations traces how the campaign, having validated its method and confirmed its deniability in the opening sequence, spent the next year doing at scale what it had spent 2021 and 2022 learning to do at all.

A reader who has followed the argument of this account to its end should carry forward one particular idea into the next link, because it is the idea that makes the next link intelligible. That idea is that escalation, when it came, was not a change of policy but a change of pace. Nothing about the campaign’s purpose, method, or strategic logic was different in the year of eliminations from what it had been in the opening sequence. What was different was confidence, and confidence had been earned. The opening sequence is best understood, in retrospect, as the period during which the campaign paid the fixed costs of becoming a campaign: the cost of building the apparatus, the cost of testing the method, the cost of discovering whether the world would impose a penalty. Those costs, once paid, did not have to be paid again. Everything that followed drew on that initial investment. The year of eliminations was expensive in its own way, but it was spared the hardest expense of all, the expense of proving that the thing could be done, because the opening sequence had already settled that question in full.

The relationship between the two links is one of cause and consequence, and it should be stated as such. The escalation of the following year was not a fresh start. It was the harvest of the opening sequence. Every reason the campaign had to accelerate, the proven method, the validated deniability, the demonstrated reach into the militant leadership, the absence of international consequence, had been established in 2021 and 2022. The next link in the chain inherited a working campaign and simply ran it harder. To understand why the killings of the following year came so fast and reached so high, the reader must first understand that the slow, careful, three-event opening had already answered every question that might otherwise have held the campaign back. The shadow war began quietly, by design and by necessity, in the nine months from June 2021 to March 2022. It did not stay quiet, and the link that follows is the story of how loud it became.

Frequently Asked Questions

Q: When did India’s shadow war against terrorists in Pakistan begin?

Publicly, the campaign became visible in mid-2021, and the most useful single date is June 23, 2021, when a car bomb detonated near Hafiz Saeed’s Lahore residence. That bomb is the first event in the opening sequence. However, the campaign as an enterprise, meaning the decision to pursue covert eliminations and the construction of the apparatus to do so, began earlier, in the roughly twenty-two-month interval after the August 2019 revocation of Article 370. The distinction matters: June 2021 is when the campaign became observable, while its true origin lies in the preparation period that produced no public events. Anyone trying to date the shadow war should be precise about whether they mean its first visible operation or its actual beginning as a strategic undertaking.

Q: What were the three founding events of the covert campaign?

Three events across nine months make up the opening sequence. The first was the June 2021 car bomb in the Johar Town area of Lahore, near a residence associated with Hafiz Saeed. The second was the January 2022 killing of Saleem Rehmani, a man on India’s list of wanted individuals, shot dead by unidentified gunmen. The third was the March 1, 2022 killing of Zahoor Mistry, an IC-814 hijacker living in Karachi under the false identity Zahid Akhund, shot at close range by two men on a motorcycle. Read together, these three events functioned as a declaration of intent, a proof of concept, and a confirmation of pattern, and they mark the moment the shadow war moved from preparation into practice.

Q: Was the Lahore car bomb the first shadow war operation?

The Lahore bomb is the first publicly visible event that this account assigns to the campaign, and in that sense it can be called the first operation. But it does not resemble the operations that followed. It was loud, it killed three bystanders, it produced rubble and televised footage, and it generated a formal accusation from Pakistan’s national security adviser. The shootings that came after were quiet, precise, and unclaimed. The bomb is best understood as a declaration of intent, a strike that reached toward the apex of the Lashkar-e-Taiba leadership and missed because the target was imprisoned, rather than as a representative example of the campaign’s mature method. Calling it the first operation is reasonable; treating it as typical is not.

Q: Who was Saleem Rehmani and why does his killing matter?

Saleem Rehmani was a man designated and sought by India who was living in Pakistan and was shot dead there in January 2022 by unidentified gunmen. Open-source information about him is limited, and he was a mid-tier figure rather than a household name. His killing matters in the chain not because of his individual prominence but because of what it demonstrated. It was the campaign’s proof of concept: a quiet test of whether a wanted man inside Pakistan could be located and eliminated without producing attributable evidence or international consequence. The test succeeded, and that success made everything that followed possible. Rehmani’s near-invisibility in the public record is consistent with the role of a proof of concept, which is observed by the people running it rather than by the public.

Q: Why was the killing of Zahoor Mistry so significant?

Zahoor Mistry was one of the five men who hijacked Indian Airlines Flight IC-814 in December 1999, and he is identified as the hijacker who fatally stabbed a young passenger, Rupin Katyal. For more than two decades he lived in Karachi under the false identity Zahid Akhund, running a furniture business, reportedly under the protection of the Pakistani intelligence apparatus. His killing in March 2022 was significant for two reasons. First, it applied the campaign’s quiet, validated method to a target the Indian public recognized and reviled, confirming that the method worked against significant figures, not only obscure ones. Second, it demonstrated that a false identity and years of protected residence no longer guaranteed safety, a message every comparable figure in Pakistan would have absorbed immediately.

Q: Did Pakistan blame India for these killings?

Yes, repeatedly. In July 2021, Pakistan’s National Security Adviser, Moeed Yusuf, held a press conference attributing the Lahore car bomb to India’s Research and Analysis Wing, stating that forensic analysis had identified an Indian national as the mastermind. Pakistan reiterated the accusation regarding the Johar Town blast in December 2022. In the case of Mistry’s killing, Pakistan again alleged Indian involvement, claiming New Delhi had hired assassins. India rejected all of these accusations. It is important to recognize that an accusation from an interested party is not the same as independent proof, and that India’s denials are equally the statements of an interested party. The accusations are part of the record, but they do not by themselves settle the question of responsibility.

Q: How did India respond to accusations about the shadow war?

India consistently denied any role in extraterritorial assassinations and characterized Pakistani allegations as propaganda. This official silence is not evidence that nothing occurred; it is the operational requirement of a covert campaign. A deniable operation only remains deniable, and therefore only remains low-risk, if its sponsor does not acknowledge it. Acknowledgment would convert a deniable act into an act of state for which the state could be formally held accountable, inviting exactly the international consequence the covert method exists to avoid. India’s denials should therefore be read as consistent with the logic of the campaign rather than as a refutation of it. The silence is what the method requires.

Q: Did intelligence preparation begin before the Lahore car bomb?

Almost certainly, and this is one of the strongest reasons to date the campaign’s true beginning earlier than June 2021. A thirty-kilogram vehicle-borne explosive placed near a specific residence in a heavily policed city presupposes detailed target knowledge, a local network capable of acquiring a vehicle and assembling explosives, and the ability to operate in Lahore without detection. None of that infrastructure assembles in the days before a bombing. It is the product of months of collection, recruitment, and logistics. Analysts who study the Indian security establishment, including the work of Saikat Datta on when the pattern was first recognized and Praveen Swami on the campaign’s initiation phase, support treating the preparation period as the real origin and the bomb as the first visible output of a process already running.

Q: Why did India choose covert killings instead of military strikes?

This choice followed from the lessons of the February 2019 Balakot airstrike. A conventional strike against Pakistan provokes an immediate conventional response and risks escalation up a ladder both nuclear-armed states have strong reasons to fear. Conventional striking is therefore a blunt instrument: too dangerous to use often and impossible to calibrate finely. A campaign meant to degrade militant infrastructure steadily over years needs a tool that can work continuously and quietly. There is also a deeper logic. The infrastructure India wished to degrade was not buildings, which are easily rebuilt, but irreplaceable people: planners, recruiters, financiers, masterminds. Reaching specific individuals where they live is a covert mission by definition, so the choice of method followed from a correct identification of the target.

Q: What is the operational signature of the shadow war killings?

The signature, established in the opening sequence and repeated for years afterward, has several consistent elements. Assailants typically arrive and depart on a motorcycle, which moves quickly through congested streets and can be abandoned without forensic cost. They wear helmets and masks, defeating closed-circuit identification in camera-saturated cities. The killing is usually a close-range shooting, often to the head, which guarantees the outcome and minimizes noise and time at the scene. No organization claims responsibility. The absence of a claim is the most strategically important feature, because it leaves the target’s organization and the host state unable to prove who was responsible, and proof is the precondition for any formal response.

Q: How many wanted men were killed in the shadow war’s first phase?

As described in this account, the opening sequence contains two confirmed killings of wanted individuals, Saleem Rehmani in January 2022 and Zahoor Mistry in March 2022, preceded by the June 2021 car bomb, which killed three bystanders but did not reach its apparent intended target. The deliberately small number is itself meaningful. It reflects an actor moving carefully along a learning curve, testing its method before betting it at scale. The contrast with the following year, when killings came in a rapid succession against more senior targets, is sharp, and that contrast is one of the clearest indications that 2021 and 2022 were an opening phase distinct in character from the escalation that followed.

Q: Was the campaign’s beginning planned or opportunistic?

The evidence points firmly toward planning rather than opportunism. Three indicators support this. First, the long preparation interval after the 2019 revocation is consistent with the deliberate construction of a covert capability from a standing start. Second, the order of the opening targets, descending from an attempt toward the apex of the leadership to a mid-tier figure and then to a notorious but organizationally peripheral hijacker, matches the behavior of an actor learning its own method on survivable targets before risking it on the hardest ones. Third, the operational signature was consistent from the start, which suggests a designed procedure rather than improvisation. A purely opportunistic campaign would show neither the patient sequencing nor the methodological consistency that the opening sequence displays.

Q: How did the killings affect militant leaders living in Pakistan?

Their effect was immediate and behavioral. The death of a hijacker who had lived safely under a false identity for more than two decades, reportedly under intelligence protection, told every comparable figure that cover and sanctuary were no longer sufficient. Senior militant figures responded by tightening their security: reducing public movement, relocating more frequently, shrinking their circles of trust, and abandoning predictable routines. This defensive crouch is itself a form of degradation, because an organization whose leadership must concentrate on personal survival has less capacity for ambitious, externally directed operations. The opening sequence began a compounding process that pressed the militant leadership harder with each passing year.

Q: Did the shadow war stop attacks against India?

No, and this is an essential qualification. The campaign imposed a continuous, deniable cost on Pakistan-based militant figures, and it degraded the leadership of the organizations involved over time. But it did not end the threat to India, as later events in this chain make painfully clear. The shadow war should be understood as one instrument among several, capable of degrading militant infrastructure and restructuring the contest between the two states, but not capable on its own of eliminating the threat. Treating the campaign as a complete solution would misread both its purpose and its results. It changed the structure of the contest; it did not end the contest.

Q: What role did the revocation of Article 370 play in starting the shadow war?

The August 2019 revocation of Jammu and Kashmir’s special status was the preceding link that created the conditions for the campaign. It effectively closed the diplomatic channel between India and Pakistan, removing any shared framework for managing or de-escalating the contest. It also hardened a conviction inside the Indian security establishment that conventional retaliation was too dangerous to use routinely and that militant infrastructure would have to be degraded at the level of the individual. With the diplomatic track closed and the conventional track judged too risky for continuous use, the covert track became the remaining option. The roughly twenty-two months between the revocation and the Lahore bomb is the interval during which the covert capability was built.

Q: Why is it difficult to prove the killings were a coordinated campaign?

Because covert operations are designed precisely to deny proof. The claim that the three opening events belong to a single campaign is an analytical inference built on pattern: shared target selection, common geography, tight timing, a consistent operational signature in the two shootings, and the absence of any claim of responsibility. That pattern is difficult to attribute to coincidence, but pattern is not proof. It remains possible in principle that the events had different sponsors or arose from internal militant rivalries. Pakistan’s attribution to India is an interested party’s accusation, not independent confirmation, and India’s denial is equally an interested party’s statement. The honest position is that the campaign interpretation is a strong, well-supported working thesis, not an established certainty.

Q: How did the shadow war change the India-Pakistan relationship?

The opening sequence inaugurated a mode of contest that had not previously existed at this scale and consistency. For decades the violence flowed predominantly from Pakistan-based organizations into India, with India responding internally or, rarely, with conventional strikes. The shadow war introduced a sustained, deniable, individually targeted counter-flow that imposed a continuous cost on Pakistani soil. This did not equalize the relationship and did not end attacks on India, but it restructured the contest. Pakistan now faced a cost it could neither prevent, nor avenge, nor even prove, and that structural change, beginning in 2021 and 2022, frames every event in the chain that follows.

Q: Why did the campaign target a hijacker from a 1999 event so many years later?

Zahoor Mistry’s killing in 2022 shows that the campaign’s target list was not limited to figures who posed an active operational threat in the present. Mistry’s significance was historical. He had hijacked IC-814 in 1999 and was identified as the man who killed a passenger, and for two decades he had lived safely in Karachi while that crime went unanswered. Targeting him served purposes beyond degrading current militant capacity. It demonstrated that time and a false identity did not confer permanent safety, a message aimed at every protected figure regardless of when his offense had occurred. It also addressed a long-standing grievance that carried real weight in Indian public memory, since the IC-814 hijacking and the concessions it forced remained a national wound. A campaign that reaches back to a figure from 1999 is signaling that its memory is long and its list does not expire, and that signal has a deterrent value independent of any single killing.

Q: What does the term shadow war mean in this context?

The phrase shadow war describes a sustained, deniable, low-visibility campaign conducted by one state against targets associated with another, without open acknowledgment and without the formal machinery of declared conflict. In this context it refers specifically to the campaign of targeted eliminations directed at wanted militant figures living inside Pakistan. The word shadow captures the two defining features of the method. The campaign operates in the shadows in the sense that it is never claimed, never officially admitted, and never formally recorded as state action, which is what preserves its deniability. It is also a war in the limited sense that it is a sustained, organized, strategically directed application of lethal force toward a coherent objective, even though it involves no armies, no front lines, and no declarations. The term is useful precisely because it names something that sits between peace and conventional war, a mode of contest that conventional vocabulary struggles to describe.

Q: What came after the shadow war’s opening phase?

The opening phase of 2021 and 2022 led directly into a year of dramatic escalation. The careful, two-killing pace of the opening sequence gave way to a rapid succession of eliminations against targets that climbed back up the ladder of organizational seniority, with the geographic spread widening beyond the Punjab heartland. The number of killings in a single year reached a level that no coincidence could explain, and that escalation made the existence of a campaign impossible for honest observers to deny. The escalation was not a new beginning; it was the harvest of the opening sequence, which had already validated the method, proven the deniability, and removed every reason for the campaign to remain cautious.