A campaign that never announced its objective also never announced an ending. India’s shadow war, the program of targeted killings of wanted terrorists on Pakistani soil that began when IC-814 hijacker Zahoor Mistry was shot dead in Karachi’s Akhtar Colony on March 1, 2022, has now run for the better part of four years without a single official acknowledgment from New Delhi, without a stated goal beyond each killing itself, and therefore without any criterion by which anyone in India, Pakistan, or Washington could stand up and declare the matter concluded. That absence is not an oversight. It is the defining feature of the campaign, and it is the reason this article exists, because the only question a covert attrition program structurally cannot answer for itself is the one that matters most for everyone living inside its blast radius: what comes next.

Prediction is an uncomfortable exercise, and it should be performed with humility. Nobody forecasting the trajectory of a deniable intelligence operation can claim certainty, because the operation’s own designers have left no public roadmap and may not themselves possess one. What follows is therefore not a forecast in the sense of a weather report. It is a structured argument about the small number of futures the shadow war can plausibly produce, the conditions under which each becomes more or less likely, and the strategic logic that pushes the campaign toward some outcomes and away from others. The honest version of this analysis acknowledges its own limits at every turn, presents probabilities rather than confident assertions, and treats the reader as an adult capable of weighing evidence rather than a consumer to be reassured.

The central claim of this analysis can be stated plainly. The shadow war has no defined endpoint because it has no defined objective beyond the elimination of India’s enemies on Pakistani soil. The campaign does not seek regime change in Islamabad. It does not seek territorial conquest. It does not seek a negotiated settlement that could be signed, ratified, and pointed to as a conclusion. It seeks the permanent degradation of Pakistan’s terror infrastructure through attrition, one funeral at a time, and a program built on attrition rather than on a political objective has only two natural destinations. Either it succeeds so completely that no targets remain, which the recruitment machinery inside Pakistan makes extraordinarily unlikely, or it continues indefinitely until it stops being a campaign at all and becomes a permanent feature of the India-Pakistan relationship, as routine and as unremarkable as the artillery exchanges that once defined the Line of Control. Everything else this article examines flows from that single structural fact.

There are four futures worth taking seriously, and the bulk of this analysis is devoted to laying each of them out with the probability, the timeline, and the consequences attached. The first future is the one optimists hope for and pessimists dismiss: Pakistan cracks down on terrorism with enough conviction that the campaign loses its justification and quietly winds down. The second future is the one the current trajectory points toward most directly: the program simply continues, year after year, as permanent low-intensity attrition. The third future is the one intelligence planners lose sleep over: a single catastrophic failure, a captured operative, a botched strike that kills the wrong people, exposes the campaign so completely that India is forced to halt it. The fourth future is the one that should worry strategists most, because it is the one where the campaign succeeds at its narrow task and then, finding no exit, escalates into something that is no longer counter-terrorism at all. Each of these deserves to be examined on its own terms, and none of them should be dismissed because it is comfortable or uncomfortable to contemplate.

A word is owed, before the scenarios begin, on why this kind of analysis is worth attempting at all, given the genuine uncertainty involved. The temptation, faced with a deniable program whose designers have published no roadmap, is either to throw up one’s hands and declare prediction impossible, or to overreach and present a single confident forecast dressed in false precision. Both responses fail the reader. The first abandons the question that matters; the second answers it dishonestly. The middle path, the one this article takes, is to accept that the campaign’s exact future is unknowable while insisting that the campaign’s possible futures are not unlimited. A covert attrition program operating between two nuclear-armed states under a no-talks doctrine is heavily constrained by its own structure, by arithmetic, and by the strategic logic of the actors involved. Those constraints do not tell us which future will arrive. They do tell us, with reasonable confidence, which futures are structurally available and which are not, and they allow each available future to be weighted by probability rather than merely listed. That is a real and useful form of knowledge, and it is the form this analysis aims to deliver.

This is also the narrative prediction that runs parallel to the broader question of whether India and Pakistan will fight another open war, and the two questions are not separable. The shadow war and the open war are not different categories of conflict. They are phases of one continuous campaign, two arms of the same strategic body, and the future of the covert program cannot be predicted without reference to the conventional confrontation that produced Operation Sindoor. What comes next in the shadows depends heavily on what comes next in the open, and the reverse is equally true.

The Preceding Link

Every forecast needs a starting position, and the shadow war’s forward trajectory cannot be plotted without first understanding the diplomatic posture that now frames it. That posture is the doctrine of no engagement, and it is the immediate predecessor of everything this article predicts.

In the months after the Pahalgam massacre of April 22, 2025, and the conventional military exchange of Operation Sindoor that followed in May, the Indian government did something that distinguished this crisis from every previous India-Pakistan confrontation since the 1972 Simla negotiations. It declined to return to the table. There was no back-channel reset, no quiet resumption of foreign-secretary contact, no composite-dialogue revival of the kind that had followed Kargil in 1999, the 2001 Parliament attack, the 26/11 Mumbai siege, and the Pathankot raid of 2016. Instead, New Delhi articulated what amounts to a permanent precondition. Prime Minister Narendra Modi’s formulation, that terror and talks cannot proceed together and that blood and water cannot flow together, was not a slogan for a single news cycle. It hardened into the diplomatic context within which the shadow war now operates: a declared zero-tolerance doctrine under which engagement resumes only when Pakistan verifiably dismantles the terror infrastructure on its soil, and not before.

The significance of this for the covert campaign is difficult to overstate. In every previous era, the shadow war, had it existed in its current form, would have been a bargaining chip. It could have been wound down as a goodwill gesture before a summit, paused during a dialogue round, traded against Pakistani concessions on Mumbai prosecutions or Kashmir infiltration. Diplomacy provided an off-ramp, a mechanism through which covert pressure could be converted into political settlement and then retired. The no-talks doctrine removes that off-ramp. If there is no dialogue, there is no negotiating table on which to place the campaign, and therefore no diplomatic process through which it could be formally concluded. The doctrine that was designed to pressure Pakistan also, as a side effect, traps the shadow war in a state of permanent activity, because it eliminates the most obvious mechanism by which covert operations have historically been terminated.

Shyam Saran, the former Indian foreign secretary, has been among the analysts most willing to ask whether the no-engagement posture is historically unprecedented, and his answer matters here. India has always re-engaged before. After every prior rupture, the underlying logic of two nuclear-armed neighbors sharing a 3,300-kilometer border eventually reasserted itself, and talks resumed, however grudgingly and however little they achieved. The question Saran raises is whether the post-Pahalgam environment is genuinely different or merely the latest iteration of a recurring pattern. If it is the latest iteration, then re-engagement will eventually arrive, and with it an opportunity to retire the covert campaign. If it is genuinely different, if the no-talks doctrine really does represent a permanent structural shift rather than an extended pause, then the shadow war loses its diplomatic exit and the prediction problem becomes far harder. This analysis treats the matter as genuinely uncertain, because the evidence cuts both ways and intellectual honesty requires saying so.

What is not uncertain is that the no-talks doctrine, the conventional confrontation of Operation Sindoor, and the ongoing covert program now form a single integrated posture. India is simultaneously refusing to talk, having demonstrated its willingness to strike conventionally across the international border, and continuing to eliminate individual targets through deniable means. These are not three separate policies. They are three expressions of one strategic decision, the decision that cross-border terrorism will no longer be a cost-free option for the Pakistani state, and that the cost will be imposed across the diplomatic, conventional, and covert domains at once. The shadow war’s future, therefore, cannot be predicted as if it were a standalone intelligence program. It must be predicted as one component of an integrated pressure strategy, and that integration is the preceding link from which every scenario in this article descends.

Where the Shadow War Stands in Early 2026

Before the futures can be assessed, the present must be measured. A forecast is only as good as its starting data, and the shadow war’s current operational profile can be described with reasonable confidence across three dimensions: tempo, target seniority, and geographic reach. Each dimension has moved in the same direction, and the direction is the most important fact about the campaign’s present state.

Begin with tempo. The early phase of the campaign, running from Mistry’s killing in March 2022 through the end of that year, was sparse. A handful of eliminations, widely spaced, easily dismissed by Pakistani authorities as the ordinary background violence of a country awash in sectarian feuds and criminal score-settling. That deniability was an asset for India and a comfort for Islamabad, because a killing every few months could be absorbed without any official having to concede that something systematic was underway. The year 2023 destroyed that comfort. The pace accelerated sharply, with a cluster of killings in September, October, and November of that year that no honest observer could attribute to coincidence. Bashir Ahmad Peer, the Hizbul Mujahideen launching commander known as Imtiyaz Alam, was shot dead outside a shop in Rawalpindi in February 2023. Paramjit Singh Panjwar, the Khalistan Commando Force chief, was killed during a morning walk in Lahore in May. The autumn brought Riyaz Ahmad in Pakistan-occupied Kashmir, Shahid Latif inside a mosque in Sialkot, Mufti Qaiser Farooq in Karachi, Dawood Malik in North Waziristan, and others, a rhythm of killings tight enough that Pakistani officials abandoned the coincidence explanation entirely and began formally accusing Indian intelligence of orchestrating the campaign. By the period following Operation Sindoor, the tempo had accelerated again, with the year 2026 producing the remaining targets being struck at a rate that observers described as unprecedented, more than thirty militants linked to Lashkar-e-Taiba, Jaish-e-Mohammed, and Hizbul Mujahideen falling inside a single calendar year. Whatever else can be said about the campaign, its operational tempo has risen monotonically since inception, and a program accelerating after four years is not a program preparing to stop.

Consider next target seniority, the second and arguably more revealing dimension. The campaign’s early victims were operationally significant but rarely famous. Couriers, launching commanders, mid-tier facilitators, men whose names meant something to intelligence analysts but nothing to the general public. Over time the seniority of the targets has climbed. Shahid Latif was a named conspirator in the 2016 Pathankot airbase attack, a man India had specifically demanded Pakistan prosecute, and his killing inside a Sialkot mosque was a deliberate statement that attack masterminds enjoy no permanent safety. The trajectory then reached its most audacious expression in the operation against Lashkar-e-Taiba co-founder Amir Hamza in Lahore, the strike examined in detail as the campaign’s climax, because reaching a co-founder, a man who had stood beside Hafiz Saeed at the organization’s creation, established that even the founding generation of Pakistan’s terror architecture could be touched. This climbing seniority is the single most important predictive indicator in the entire analysis. A campaign that starts with couriers and works its way up toward co-founders is a campaign moving up an organizational ladder, and the question of what happens when it reaches the top of that ladder is precisely the question the four-scenario framework exists to answer.

The third dimension is geographic reach, and here the pattern is one of steady expansion from the periphery toward the core. The earliest killings occurred in places where deniability was easiest to sustain, in Karachi’s sprawling anonymity and in the contested ambiguity of Pakistan-occupied Kashmir. Over time the strikes moved inward and upward, into Rawalpindi, the garrison city that houses the headquarters of the Pakistan Army, and into Lahore, the political and cultural capital of Punjab and the heartland of the Pakistani establishment. A killing in Karachi can be explained away. A killing in Rawalpindi, within reach of General Headquarters, cannot, because it carries an unmistakable message about the limits of the protection the Pakistani state can offer even to those it most wants to shelter. The geographic trajectory, like the tempo and the seniority, points inward toward the Pakistani core rather than outward toward the margins.

Three dimensions, one direction. Tempo rising, seniority climbing, geography moving toward the center. The detailed target assessment of who remains on India’s list and the comprehensive ledger of those already eliminated together confirm that this is not a campaign in decline, not a campaign reaching natural exhaustion, and not a campaign whose designers are visibly preparing to declare victory and stand down. It is a program in its most intense phase to date. Any honest forecast must begin from that starting position, and any scenario that imagines the campaign simply fading away must explain how a program accelerating on all three measured dimensions would reverse itself. That is the analytical floor beneath everything that follows.

One further feature of the present state deserves emphasis, because it bears directly on every scenario that follows, and that is the operational model through which the campaign is conducted. The shadow war is not, on the available evidence, a program of Indian operatives moving through Pakistani cities. It is a program built on intermediaries. The reporting that has emerged, including The Guardian’s 2024 account citing Indian intelligence officers, describes a model in which cells recruit and pay local criminals, and in some accounts local jihadists, to perform the actual killings, with the chain of tasking and payment kept deliberately long and deliberately deniable. This design is the source of the campaign’s central strength and its central fragility at once. The strength is that no Indian citizen is exposed at the point of the trigger, which keeps both the financial cost and the diplomatic blast radius of any single failure contained. The fragility is that the perimeter is staffed by people with no ideological stake in India’s success, people who were hired and who can therefore be caught, outbid, or turned. A program of this design can run cheaply and quietly for a very long time, which favors the second future, but it also carries a permanent and irreducible exposure risk, which keeps the third future alive. The operational model, in other words, does not merely describe how the campaign works. It pre-shapes which futures the campaign can produce.

It is worth pausing on what this model implies for the campaign’s sustainability, because sustainability is the hinge on which the second future turns. A program that depended on a finite stock of specialized Indian personnel would face natural limits, the limits of training, of rotation, of the cumulative risk to a small cadre of irreplaceable people. The intermediary model removes those limits. The supply of criminals and opportunists willing to perform a killing for money is, in a country of Pakistan’s size and economic stress, effectively inexhaustible, and each intermediary is individually disposable from the campaign’s point of view. This is precisely the property that makes indefinite continuation so structurally probable. A program is sustainable for as long as its key input is cheap and abundant, and the shadow war’s key input, deniable local muscle, is both. The campaign could, on the logic of its own design, continue at its current tempo for a decade or longer without straining any resource that India would find difficult to replace. Sustainability is not a constraint on this program. It is one of its defining features, and any forecast that imagines the campaign ending through simple exhaustion has misunderstood how the campaign is built.

The First Future Pakistan Chooses to Dismantle

The first plausible future is the one in which the shadow war ends because it succeeds at the level of strategic effect rather than at the level of body count. In this scenario, the Pakistani state concludes that the cost of sheltering anti-India militancy has finally exceeded the benefit, and it moves, genuinely and verifiably, to dismantle the infrastructure that the campaign was designed to punish. The justification for the killings evaporates, India has no further targets it considers worth the operational risk, and the program winds down not through any treaty or announcement but through the simple disappearance of its rationale.

This is the outcome that maps most directly onto what the counter-terrorism scholar Audrey Kurth Cronin, in her study of how terrorist campaigns reach their conclusions, would recognize as a particular kind of ending. Cronin’s central insight, drawn from two centuries of cases, is that terrorist campaigns always end, and that they end through a small number of identifiable pathways: the removal of leadership, negotiation, the achievement of goals, internal collapse, repression, and reorientation toward other forms of violence. The first future described here is essentially a wager that sustained pressure, the covert killings combined with the conventional shock of Operation Sindoor and the diplomatic and economic isolation of the no-talks doctrine, can push Pakistan’s terror ecosystem down the pathway of strategic abandonment, where the sponsoring state simply decides the instrument is no longer worth maintaining.

There are reasons this future is not pure fantasy. The Pakistani state’s calculus around militant proxies has always been a cost-benefit calculation rather than an ideological commitment, and the costs have risen dramatically. Operation Sindoor demonstrated that India would now strike terrorist infrastructure deep inside Pakistani Punjab, not merely in Pakistan-occupied Kashmir, and that the Indian political system would absorb the escalation risk of doing so. The economic dimension has bitten as well, with trade frozen and the broader machinery of bilateral exchange dismantled. The Financial Action Task Force process, the international scrutiny of terror financing, and the simple reputational cost of being repeatedly named as the epicenter of regional terrorism all push in the same direction. A Pakistani establishment performing an honest ledger might conclude that the proxy strategy, conceived decades ago as a low-cost method of bleeding a larger neighbor, has become a high-cost strategy that invites missiles onto Pakistani soil and isolates the country diplomatically while delivering diminishing strategic returns.

And yet the probability that must be assigned to this future is low, and intellectual honesty requires explaining why rather than simply asserting the pessimistic case. The obstacle is structural, and it sits inside the Pakistani state itself. The militant infrastructure is not a foreign body that the establishment could excise if it chose. It is woven into the institutional fabric, into the doctrine of the military establishment that has, across decades, treated groups like Lashkar-e-Taiba and Jaish-e-Mohammed as instruments of strategic depth and as low-cost tools for the Kashmir dispute. Dismantling that infrastructure would require the Pakistan Army to act against networks it built, sheltered, and in some accounts directed, and to do so it would have to overcome not only institutional inertia but the genuine risk that demobilized militants would turn their weapons inward, as Pakistan’s own experience with the Tehreek-e-Taliban Pakistan has repeatedly demonstrated. A crackdown is not a policy lever that Islamabad can simply pull. It is a wrenching internal transformation that would pit the state against parts of itself, and states rarely undertake such transformations except under existential pressure.

There is a further complication that makes this future even less likely in the near term. Army Chief Asim Munir, elevated to the rank of Field Marshal during the 2025 crisis, has built his political standing in part on a posture of defiance toward India, articulated in rhetoric that framed Kashmir in the language of existential national identity. A military leadership that has invested its prestige in defiance cannot easily pivot to the visible concession that a genuine crackdown would represent, because the pivot would look like capitulation to Indian coercion, and capitulation is precisely what an establishment built on the two-nation theory and on parity with a larger neighbor cannot afford to be seen performing. The internal politics of the Pakistani state, in other words, make the first future hard to reach not because Pakistani leaders are irrational but because the rational move for any individual leader is to avoid being the one who blinked.

If the first future arrives at all, it is most likely to arrive slowly, partially, and deniably, the way the doctrinal prediction of India’s evolving posture anticipates. Not a dramatic announcement of dismantlement but a quiet attenuation: fewer training camps operating openly, recruitment networks pushed further underground, the most internationally toxic figures eased out of visible roles. A partial and deniable Pakistani drawdown would create a genuine dilemma for India, because the no-talks doctrine demands verifiable dismantlement, and verification of a deliberately deniable process is close to impossible. The first future, even in its most optimistic version, therefore does not produce a clean ending. It produces an ambiguous situation in which India must decide whether quiet, unverifiable, partial Pakistani restraint is enough to justify standing the campaign down, and that decision returns the analysis to the central problem of a campaign with no defined criterion for success. Probability assigned to this future as a clean, campaign-ending outcome within five years: low, perhaps one in six. Probability of a partial, ambiguous version that complicates rather than concludes the campaign: meaningfully higher, and itself a reason the shadow war is so difficult to predict.

History offers a small number of cases in which a state genuinely abandoned a proxy strategy, and they are worth examining because they reveal what the first future would actually require. In Cronin’s account, the clearest examples of the success or strategic-abandonment pathway tend to coincide with large structural shifts rather than with the steady application of pressure alone. Sponsors abandoned proxies when the geopolitical context that made the proxies useful disappeared, when the end of the Cold War removed the patron-client logic that had sustained a particular insurgency, or when a sponsoring state’s own survival came to depend on the change. The pattern is instructive. It suggests that the first future is unlikely to arrive as a direct response to the shadow war’s pressure, because pressure alone has rarely been sufficient. It is more likely to arrive, if it arrives at all, as a byproduct of some larger transformation in Pakistan’s strategic situation, a shift in its relationship with China, a shift in its economic dependence on international institutions, or a shift in the internal balance between its civilian and military power centers.

The international financial dimension is the most plausible vector for such a transformation, and it deserves a fair assessment. Pakistan’s repeated appearances on the Financial Action Task Force grey list, and the conditions attached to its international lending arrangements, have already forced measurable changes in how openly terror financing can operate. A Pakistani establishment facing genuine economic distress, dependent on the goodwill of institutions that demand visible action against terror financing, has a material incentive to constrain the most internationally toxic parts of the militant ecosystem. This is not the same as a crackdown, and it should not be confused with one. But it is a channel through which external pressure unrelated to the shadow war could, over years, erode the infrastructure the campaign targets, and if the first future arrives, this slow financial strangulation is a more probable mechanism than any dramatic Pakistani change of heart. The implication for prediction is sobering for optimists. The cleanest ending for the shadow war may depend less on anything India does in the shadows than on forces operating in the open ledgers of international finance, on a timescale measured in many years rather than in a single decisive season.

The Second Future the Campaign Without End

The second future requires no dramatic event, no Pakistani transformation, no catastrophic failure. It requires only that the present trajectory continue, and for that reason it is the future this analysis judges most probable. In this scenario, the shadow war neither succeeds in the sense of provoking a Pakistani crackdown nor fails in the sense of being exposed and halted. It simply persists. The killings continue at roughly the current tempo, the targets are replenished as fast as they are eliminated, and the campaign settles into a steady state that no longer resembles a campaign at all because campaigns, by definition, are bounded efforts with beginnings and ends. What the second future describes is the conversion of the shadow war from a campaign into a condition.

The strategic logic that makes this future the default is worth tracing carefully. A campaign continues as long as the actor conducting it judges the benefits to exceed the costs, and for India the cost-benefit ratio of the shadow war is, at present, favorable. The operational costs are low, because the campaign relies on locally recruited proxies and criminal intermediaries rather than on Indian personnel exposed inside Pakistan, which limits both the financial expense and the diplomatic damage of any single failure. The deniability, while eroding, still functions, because no operation has produced a confession or a captured Indian operative that would convert allegation into proof. The deterrent benefit is real, because every killing reinforces the message that Pakistani soil offers no permanent safety. And the domestic political benefit, in an Indian environment where decisive action against cross-border terrorism enjoys broad public approval, is substantial. When the benefits are real and the costs are contained, the rational decision is to continue, and the second future is simply the projection of that rational decision indefinitely forward.

Cronin’s framework illuminates why this steady-state outcome is so common and so durable. Among her pathways, repression, the attempt to end a campaign of violence through sustained force, is the most frequently chosen and, she argues, often among the least conclusive, because force can suppress the visible manifestations of a network without resolving the conditions that regenerate it. The shadow war is, in essence, a repression strategy conducted by covert means against the personnel of Pakistan’s terror infrastructure. If Cronin’s historical analysis holds, such a strategy is likely to produce exactly the steady state the second future describes: continuous activity that suppresses without concluding, that manages a problem without solving it. The campaign becomes, in this reading, a permanent maintenance operation, the strategic equivalent of mowing a lawn that will not stop growing.

George Perkovich, the Carnegie scholar who has written extensively on the strategic dynamics between India and Pakistan, has framed the deeper question that the second future raises: whether permanent attrition is a viable strategic posture between two nuclear-armed states, or whether it is a posture that merely defers a reckoning. The viability case rests on the observation that the second future is, in cold strategic terms, stable. It imposes ongoing costs on Pakistan, it satisfies Indian domestic demand for action, it stays below the threshold that would compel a conventional Pakistani military response, and it requires no risky political decisions by anyone. A posture that is cheap, popular, and stable is a posture that tends to persist. The counter-case, which Perkovich’s broader work takes seriously, is that permanent attrition is stable only until it is not, that a posture relying indefinitely on deniability and on staying below thresholds is a posture accumulating risk, and that the second future is less a stable equilibrium than a long fuse.

What the second future looks like from the inside, lived rather than analyzed, is a normalization of the abnormal. A killing every few weeks somewhere in Pakistan, briefly noted and quickly forgotten. Pakistani officials issuing routine condemnations that no longer carry the charge of novelty. Indian officials maintaining the practiced silence that has characterized the campaign from the start. International coverage thinning as the killings lose the quality of news. The shadow war recedes into the background hum of the South Asian security environment, joining the LoC ceasefire violations, the periodic infiltration attempts, and the recurring diplomatic frost as one more permanent irritant in a relationship defined by permanent irritants. This is the future in which the campaign does not end so much as disappear into the texture of normal, and it is the future toward which, absent some disrupting event, the current trajectory most directly points. Probability assigned: the highest of the four, on the order of two in five, precisely because it is the future that requires nothing to happen.

There is a cost to the second future that the cold strategic accounting tends to understate, and intellectual honesty requires naming it. A campaign that becomes a permanent condition does not merely persist. It normalizes a category of state behavior that, when it began, was treated as exceptional. The first killings of the shadow war were extraordinary events, each one a small crisis, each one capable of generating headlines and diplomatic exchanges. The second future is the future in which that extraordinariness is worn away by repetition, until the targeted killing of a human being in a foreign city becomes an unremarkable administrative routine, processed by the system and forgotten by the public. Cronin’s framework includes the audience as one of the three actors whose interaction determines how a campaign ends, and the second future is, in part, a story about the audience: about an Indian public that grows accustomed to the program, a Pakistani public that absorbs the killings into the background of national grievance, and an international audience whose attention moves elsewhere. Normalization is the second future’s quiet engine, and it is also its quiet cost, because a state that has normalized covert killing as a routine instrument has changed something about itself that is not easily changed back.

The historical analogy that fits the second future most closely is not another counter-terrorism campaign but the long, low-intensity confrontations that have defined other enduring rivalries, the kind of permanent friction that neither escalates to open war nor resolves into peace. Such confrontations can persist for decades, and they tend to persist precisely because they are survivable, because each side learns to absorb the costs the other imposes, and because the absence of catastrophe is mistaken for the presence of stability. The second future is the shadow war settling into that pattern, becoming one more permanent feature of a relationship already crowded with permanent features. It is survivable. It is, on current evidence, the most probable outcome. And it is worth saying plainly that survivable and probable are not the same as good, because a permanent covert killing program is a heavy thing for a state to carry indefinitely, and the weight does not disappear merely because the carrying has become routine.

The Third Future Exposure and Collapse

The third future is the one that intelligence planners are paid to worry about, and it is the future in which the shadow war ends not because it succeeds and not because it persists but because it suffers a failure catastrophic enough to make continuation politically impossible. This is the pathway of exposure, and it deserves careful examination both because its consequences would be severe and because the campaign’s own design, the very features that make it cheap and deniable, also make it vulnerable to this particular kind of collapse.

To understand how the third future arrives, it helps to be specific about what catastrophic failure would actually look like, because the abstraction “exposure” conceals several distinct failure modes. The first and most damaging would be a captured operative who talks. The campaign’s reliance on locally recruited proxies and criminal intermediaries, the feature that protects Indian personnel and limits direct exposure, also means the campaign’s perimeter is staffed by individuals with no ideological loyalty to India, individuals who were paid and who could therefore be caught, flipped, and paraded. A Pakistani counter-intelligence success that produced a credible intermediary, on camera, describing the chain of payment and tasking that led back toward Indian intelligence would convert four years of allegation into something close to proof, and proof changes everything. The second failure mode is the botched operation that kills the wrong people, a strike on a target that turns out to have been misidentified, or an operation whose collateral casualties include children or other unambiguous innocents. The shadow war’s legitimacy, such as it is, rests entirely on the claim that its victims are terrorists. A killing that visibly violated that claim would strip the campaign of the moral framing that has insulated it from sustained international criticism. The third failure mode is the operation that crosses an unintended line, a strike that kills someone whose death triggers a response the campaign’s designers did not anticipate, and that mode shades directly into the fourth future examined below.

The Guardian’s 2024 reporting, which cited Indian intelligence officers describing the alleged program and presented documentation said to detail direct involvement by India’s external intelligence service, demonstrated that the third future is not hypothetical but is already partially underway. That reporting did not collapse the campaign, because journalism, however credible, is not the same as a confession or a captured operative, and India was able to dismiss it as propaganda. But it established the template. It showed that the campaign’s secrecy is not absolute, that insiders can and do talk, and that the gap between the present state of plausible deniability and the third future’s state of undeniable exposure is narrower than the campaign’s continued operation might suggest. Each additional killing is an additional opportunity for the failure modes above to occur, and probability, compounded across hundreds of operations, accumulates.

What makes the third future genuinely dangerous, as opposed to merely embarrassing, is the interaction between exposure and the post-Sindoor strategic environment. Before 2025, exposure of the shadow war would have been a diplomatic problem, costly to India’s international standing but absorbable. After Sindoor, the calculus is different, because the conventional threshold has already been crossed once. A catastrophic exposure of the covert campaign would arrive in an environment where India and Pakistan have demonstrated, recently and to each other, a willingness to exchange missile fire across the international border. Exposure that humiliated the Pakistani military, by proving that India had been killing people inside Pakistan with impunity for years, could generate domestic Pakistani pressure for a response, and the response menu in the post-Sindoor world includes conventional options that would not have been on the table a decade earlier. The third future, in its most dangerous form, is not simply the end of the shadow war. It is the end of the shadow war followed by a conventional crisis triggered by the manner of its ending.

There is also a quieter version of the third future, and honesty requires presenting it alongside the dramatic version. In the quieter version, exposure is gradual rather than sudden, an accumulation of credible reporting, leaked documents, and partial confessions that never reaches a single catastrophic moment but that, over time, makes the campaign too internationally costly to sustain. India’s relationships with Western partners, several of whom have themselves accused Indian intelligence of extraterritorial operations on their own soil, impose a ceiling on how much exposure New Delhi can absorb before the diplomatic price of the campaign exceeds its strategic value. The quieter third future is the campaign being wound down not because of one disaster but because the cumulative weight of exposure tips the cost-benefit calculation that the second future depends upon. Probability assigned to the third future in some form, dramatic or quiet, within the next several years: moderate, perhaps one in five, with the quiet version more likely than the catastrophic one, but the catastrophic version carrying consequences severe enough to demand attention regardless of its probability.

The international dimension of the third future deserves fuller treatment, because it is the channel through which the quiet version of exposure is most likely to operate, and because it connects the shadow war to a broader pattern that India cannot easily wish away. In recent years, allegations of Indian extraterritorial operations have not been confined to Pakistan. Partners and friends of India, states with which New Delhi maintains close and valuable relationships, have themselves raised accusations of Indian intelligence involvement in operations on their own soil. The significance of this for the shadow war is indirect but real. It means the campaign in Pakistan does not exist in diplomatic isolation. It is read, by India’s Western partners, as one instance of a wider willingness to operate covertly across borders, and that wider reading raises the diplomatic cost of every individual revelation. A single exposed killing in Pakistan, taken alone, might be absorbable. The same exposure, taken as confirmation of a pattern that has already strained relationships with capitals India values far more than Islamabad, is considerably harder to absorb.

This creates a ceiling, and the ceiling is the most important feature of the quiet third future. India’s strategic position depends heavily on its partnerships with the world’s major democracies, on technology access, on defense cooperation, and on the diplomatic alignment that helps constrain China. The shadow war, however satisfying domestically and however effective tactically, is a program those partnerships do not formally endorse and that, if exposed too completely, those partners would find embarrassing to be associated with. There is therefore a level of exposure beyond which the campaign’s diplomatic cost begins to threaten relationships that matter more to India’s long-term security than the elimination of any individual militant. The quiet third future is the program being wound down not because of one televised confession but because the cumulative weight of reporting, leaks, and partial confirmations carries it past that ceiling. It is a slower and less dramatic ending than the catastrophic version, but it is also, for that reason, the more probable of the two, and it is an ending driven less by what happens inside Pakistan than by what India’s most important friends are willing to tolerate.

The Fourth Future Escalation Past Counter-Terror

The fourth future is the most unsettling, and it is unsettling precisely because it is not a failure scenario. It is a success scenario that does not stop. In this future, the shadow war works. It climbs the organizational ladder, eliminates the operational figures, reaches the senior commanders, and approaches the top of the terror hierarchy. And then, having succeeded at its narrow task and finding no exit, it does not conclude. It escalates, expanding its definition of a legitimate target until the campaign is no longer recognizably counter-terrorism and has become something else, something with no name in the current vocabulary because states do not usually admit to conducting it.

The logic of the fourth future is the logic of a campaign with no exit criteria meeting a campaign that has run out of the targets it was designed to hit. Recall the central claim of this analysis: the shadow war has no defined endpoint because it has no defined objective beyond elimination on Pakistani soil. Now follow that claim forward. Suppose the second future does not materialize because the campaign is more successful than steady-state attrition implies. Suppose the climbing seniority documented earlier continues until the conventional terror leadership, the co-founders and the operational chiefs, have been substantially degraded. A campaign with a political objective would, at that point, declare victory and stop. A campaign without one faces a choice: stop anyway, despite having no criterion that says it is time, or continue by redefining what counts as a target. The fourth future is the future in which the campaign chooses to continue, and the only way to continue once the obvious targets are exhausted is to expand the target set.

Expansion has a direction, and the direction is toward the state. The terror infrastructure that the shadow war targets does not float free of the Pakistani establishment. It is, in India’s own consistent analysis, sponsored, protected, and in some measure directed by elements of the Pakistani military and intelligence services. Once the campaign has worked through the militants themselves, the network it is mapping points unavoidably toward the state officials who enabled those militants. The fourth future is the future in which the campaign follows that pointer, in which an operation eliminates not a Lashkar commander but a serving or retired intelligence officer who handled him, not a Jaish facilitator but the military figure who protected the facilitator’s seminary. At that moment the campaign changes character. Killing a designated terrorist, however legally contested, can be framed as counter-terrorism. Killing a serving officer of another state’s armed forces cannot. It is an act of war by any definition the international system recognizes, conducted covertly, and a covert act of war between nuclear-armed neighbors is a category of event with almost no stabilizing precedent.

It is important to be clear about what is being claimed here, because the fourth future is the most speculative of the four and the easiest to overstate. There is no evidence that India has decided to escalate the campaign toward state targets, and the analysis is not asserting that this future is probable. It is asserting that this future is structurally available, that the absence of exit criteria creates a pressure toward it, and that the pressure is real even if no one has yet acted on it. A campaign designed to stop somewhere has a built-in brake. A campaign designed only to continue eliminating has, instead, a built-in accelerator, and the fourth future is what happens when the accelerator is followed past the boundary that separates counter-terrorism from interstate covert warfare. The danger is not that India’s leadership wants this outcome. The danger is that a campaign without a defined endpoint can drift toward it incrementally, each individual target only slightly more senior than the last, until the cumulative drift has carried the campaign across a line that no single decision ever consciously chose to cross.

The consequences of the fourth future are the most severe of any scenario in this article, because they reintroduce, in a particularly unstable form, the escalation dynamics that Operation Sindoor already demonstrated. A conventional crisis, like Sindoor, has at least the virtue of legibility: both sides can see the strikes, count the costs, and signal through established channels, and the 2025 crisis did terminate in a director-general-level ceasefire precisely because the conventional domain retains mechanisms for controlled de-escalation. A covert campaign that has escalated to killing state officials offers no such legibility. The attacks are deniable, so they cannot be openly acknowledged or openly negotiated over. The targets are state personnel, so the offense is grave enough to demand a response. And the response, lacking a covert ceasefire mechanism, may have to be conventional, which means the fourth future’s most plausible terminus is the open war that the shadow war was, in part, supposed to make unnecessary. Probability assigned: the lowest of the four, perhaps one in eight, but weighted in any rational assessment by the severity of its consequences, because a low-probability path to interstate war between nuclear powers is not a path that strategic planning can afford to ignore.

It is worth dwelling on why the fourth future is so much more dangerous than the others, because the danger is not obvious and the explanation reveals something important about the entire program. Covert killing between states is not unprecedented. The historical record contains episodes in which intelligence services killed, or attempted to kill, the officials and agents of rival states. But the record also shows that such episodes are most dangerous, and most likely to spiral, precisely when they are deniable, because deniability removes the mechanisms through which states de-escalate. When a confrontation is open, as Operation Sindoor was open, both governments can see what is happening, can communicate about it through established channels, and can negotiate a halt, which is exactly how the 2025 crisis terminated, in a conversation between the two directors-general of military operations. A covert campaign offers no equivalent. There is no covert ceasefire mechanism, no deniable version of a hotline, no way for two governments to formally agree to stop doing something that neither will admit to doing. The fourth future is dangerous because it would combine the gravity of attacks on state personnel with the ungovernability of the covert domain, and that combination has no built-in brake.

There is a second reason the fourth future merits attention out of proportion to its probability, and it concerns the nature of incremental drift. The other three futures each require a recognizable event or decision: a Pakistani crackdown, a sustained continuation, a catastrophic exposure. The fourth requires none. It can arrive through the accumulation of individually defensible choices, each target only marginally more senior than the last, each operation justified on the same counter-terrorism logic as the one before it, until the campaign has crossed a line that no single decision was ever framed as crossing. This is the characteristic danger of any program that lacks a defined endpoint. The absence of a stated limit means the absence of a moment at which someone must consciously decide to go further, and decisions that are never consciously made are decisions that escape the scrutiny that might otherwise restrain them. The fourth future is not a prediction that India intends interstate covert warfare. It is a warning that a campaign built without an exit has removed the very checkpoint at which such an escalation would have to be deliberately chosen, and that removal is itself the hazard.

A Campaign Built Without an Exit

The four futures share a single root, and that root is worth excavating directly, because it is the deepest structural fact about the shadow war and the fact that makes the campaign so resistant to prediction. The shadow war was built without an exit. It has no stated objective whose achievement would signal completion, no metric whose satisfaction would justify a stand-down, and no mechanism, diplomatic or operational, designed to conclude it. This section examines that absence as a strategic question in its own right, and it adjudicates the central disagreement the question provokes: whether a campaign without exit criteria can achieve strategic success, or whether the absence of an endpoint is itself a form of strategic failure.

Begin with the case that the absence of an endpoint is a feature rather than a flaw, because that case is stronger than it first appears and deserves a fair hearing. A campaign with a declared objective is a campaign that can fail visibly, because a declared objective is a standard against which performance is measured, and failure to meet the standard is a defeat that adversaries can point to and domestic critics can exploit. A campaign with no declared objective cannot fail in that sense, because there is no benchmark to fall short of. The shadow war’s deliberate objectivelessness, on this reading, is a sophisticated design choice. It makes the campaign immune to the charge of failure, it preserves maximum flexibility for the operators, it denies Pakistan the satisfaction of ever being able to say that India tried something and it did not work, and it keeps the adversary permanently uncertain, because an adversary who does not know what would make India stop cannot calculate what concession would buy relief. Ambiguity, in this defense, is a weapon, and the campaign without an exit is ambiguity weaponized.

The case on the other side is that this defense confuses the avoidance of visible failure with the achievement of success, and that the two are not the same thing. A campaign that cannot fail because it has no objective also cannot succeed, for exactly the same reason. Success is the achievement of an objective, and a program that has declined to name an objective has, by construction, placed success beyond its own reach. What the shadow war can achieve, on this reading, is activity, the continuous imposition of costs on Pakistan, and the maintenance of a deterrent posture. But activity is not the same as victory. The campaign can run forever and impose costs forever without ever arriving at a state of affairs that anyone could describe as having won, and a campaign that can run forever without winning is a campaign that has, in the most important sense, no strategy at all, only a tactic repeated indefinitely. Cronin’s study of how terrorist and counter-terrorist campaigns end is, at bottom, a study of the importance of having a theory of the ending, and her work suggests that campaigns conducted without such a theory tend to outlive their own usefulness, consuming resources and generating risks long after the conditions that justified them have changed.

The adjudication this analysis offers is that both positions are partly right, and that the resolution lies in distinguishing between the campaign’s tactical design and its strategic design. As a tactical instrument, the shadow war’s objectivelessness is genuinely clever, and the defenders of the design are correct that ambiguity has imposed real costs on Pakistan and has preserved real flexibility for India. As a strategic posture, however, the absence of an exit is a genuine vulnerability, and the critics are correct that a campaign that cannot define its own success has outsourced the question of its ending to chance, to Pakistani behavior, and to the accumulation of operational risk. The honest verdict is that the shadow war is a tactically excellent instrument attached to a strategically incomplete plan. It does what it does extremely well. What it does not do is contain within itself any account of when, why, or how it should stop, and that omission is not a small one. It is the omission that makes all four of this article’s futures possible, because a campaign with a defined exit would foreclose the second future of endless drift and the fourth future of escalation past counter-terrorism, leaving only the cleaner endings of success or exposure.

Perkovich’s framing of permanent attrition as a strategic question between nuclear states sharpens the verdict further. Between conventional powers, a campaign without an exit is a problem of waste and risk. Between nuclear powers, it is a problem of waste and risk and existential hazard, because the accumulation of risk in a nuclearized relationship is the accumulation of a particular and catastrophic kind of risk. The shadow war’s missing exit is therefore not merely an untidiness in the planning. It is a structural feature that loads the India-Pakistan relationship with an open-ended source of escalation potential, and the fact that the loading has not yet detonated is not evidence that it is safe. It is evidence only that it has not detonated yet. A responsible strategic assessment would treat the construction of a campaign without an exit as a decision that ought to be revisited, not because the campaign has failed, but because a campaign that cannot say how it ends has left its own ending to be written by events rather than by intent. And events, in the India-Pakistan relationship, have rarely been gentle authors.

One more analytical lens sharpens the verdict on the missing exit, and it is the lens Cronin places at the foundation of her entire study: the insistence that a campaign of violence cannot be understood by examining only the actor conducting it, but must be understood as a dynamic interaction among three parties, the group, the government, and the audience. Applied to the shadow war, the triad is revealing. The first party is the network of militant organizations the campaign targets. The second is the Indian state conducting the program and the Pakistani state on whose soil it operates. The third, the audience, is everyone watching: the Indian public, the Pakistani public, the diaspora communities, the international observers, and the future recruits weighing whether a militant career is survivable. Cronin’s central methodological claim is that a campaign reaches its conclusion only when the interaction among all three parties shifts, and the shadow war’s missing exit looks different, and worse, when viewed through that claim.

Consider the audience dimension specifically, because it is the one the campaign’s designers can least control and the one that most clearly exposes the limits of an objectiveless program. The shadow war’s implicit theory of the audience is a deterrence theory: that each killing teaches the watching pool of potential recruits that Pakistani soil offers no permanent safety, and that this lesson, accumulated, suppresses recruitment. But deterrence through fear is a notoriously unstable instrument against an audience that includes the ideologically committed, because for some members of that audience the demonstration that India is hunting them is not a deterrent but a confirmation, a proof of the narrative of existential struggle that drew them toward militancy in the first place. The campaign may deter the opportunist while inadvertently validating the believer, and a program with no defined endpoint has no mechanism for noticing when this is happening or for adjusting when the audience effect turns counterproductive. A campaign with an objective could measure progress against that objective and detect when the audience was responding perversely. A program defined only by continuous elimination has no such instrument. It cannot read its own audience, and it therefore cannot know whether it is winning the only contest, the contest for the next generation of recruits, that ultimately determines whether the threat persists.

This is the deepest sense in which the absence of an exit is a strategic failure rather than a clever design. An exit is not merely a stopping point. It is a definition of success, and a definition of success is what allows a campaign to know whether it is working. Without one, the shadow war is flying without instruments. It can see its immediate output, the count of eliminations, with perfect clarity. It cannot see its actual effect on the three-party system that determines whether the conflict is moving toward an ending or merely being fed. The missing exit, in the end, is not a gap in the paperwork. It is a gap in the program’s ability to perceive its own consequences, and a campaign that cannot perceive its own consequences is a campaign whose ending will be written, as this section has argued, by events rather than by intent.

Can Attrition Outrun the Madrassa Pipeline

Every scenario in this article ultimately depends on a single empirical question, and the question is arithmetic before it is anything else. Can the shadow war eliminate Pakistan-based militants faster than Pakistan’s recruitment and training infrastructure can produce replacements? If the answer is yes, the first future of strategic success becomes reachable, because a campaign that outruns the pipeline can, in principle, drain the pool. If the answer is no, the first future is closed, the second future of permanent attrition becomes close to inevitable, and the campaign is revealed as a program that can manage the threat indefinitely but can never eliminate it. This section examines that arithmetic, and the examination is not encouraging for anyone hoping the campaign has a clean ending.

The recruitment infrastructure in question is often referred to, with some imprecision, as the madrassa pipeline, and the imprecision is worth correcting before the analysis proceeds. Pakistan contains tens of thousands of religious seminaries, and the overwhelming majority of them are exactly what they appear to be, institutions of religious education with no connection to militancy. The pipeline that matters for this analysis is a far smaller subset, the specific network of seminaries, recruitment cells, and training facilities associated with organizations like Lashkar-e-Taiba and Jaish-e-Mohammed, institutions like the Markaz at Muridke and the Bahawalpur complex that Operation Sindoor’s strikes were directed against. It is this subset, not the broad seminary system, that constitutes the production capacity the shadow war is racing against. But even confined to that subset, the production capacity is substantial, because the militant recruitment system is not merely a set of buildings. It is a social process, embedded in particular districts of Punjab and elsewhere, drawing on networks of family, mosque, and community, sustained by an ideological apparatus and by the patronage of state elements, and capable of identifying, indoctrinating, and training new personnel on a timescale measured in months.

Set the arithmetic out honestly. The shadow war, even in its accelerated 2026 phase, eliminates militants at a rate of a few dozen per year. The recruitment pipeline, even on conservative estimates, can generate new foot soldiers and mid-tier operatives at a rate that comfortably exceeds that figure, because indoctrination and basic training are not bottlenecked the way the production of senior leadership is. This asymmetry is the single most important fact in the section, and it points to a crucial distinction that the rest of the analysis must respect. The pipeline replaces foot soldiers and operatives quickly. It cannot replace senior leadership quickly, because a co-founder, an experienced operational commander, a figure with decades of network knowledge and personal relationships across the militant ecosystem, is the product not of a months-long training cycle but of a career, and careers cannot be mass-produced. The shadow war, therefore, cannot outrun the pipeline at the bottom of the organizational ladder. It can plausibly outrun the pipeline at the top.

This distinction reframes the entire question of strategic success and connects directly to the campaign’s documented behavior. Recall that the seniority of the targets has climbed steadily, from couriers toward co-founders. That climb is not an accident and not merely a demonstration of capability. It may be the campaign’s implicit answer to the pipeline arithmetic. If the bottom of the ladder cannot be drained because the pipeline refills it faster than the campaign can empty it, then the only level at which attrition can achieve a durable effect is the top, where each elimination removes irreplaceable human capital that the pipeline cannot quickly regenerate. A shadow war that understood its own arithmetic would do precisely what the actual shadow war appears to be doing: concentrate on senior leadership, accept that the foot-soldier layer is beyond its reach, and aim not to empty the pool but to decapitate the organizations that give the pool its direction. Cronin’s pathway of decapitation, the removal of leadership, is relevant here, though her own assessment of it is cautious, because the historical record on decapitation is mixed, and organizations with deep benches or with state sponsorship have often proven able to regenerate leadership over time.

Which leads to the hard conclusion of this section, the conclusion that constrains every optimistic scenario in the article. Attrition cannot outrun the madrassa pipeline at the level of personnel, and any expectation that the shadow war will simply run Pakistan’s terror infrastructure out of people is an expectation that the arithmetic does not support. What attrition can plausibly do is degrade the leadership layer faster than that layer regenerates, and if the campaign sustains its current focus on senior figures, it can impose a real and lasting cost on the coherence and operational sophistication of the targeted organizations. But degrading leadership is not the same as eliminating the threat, because as long as the pipeline keeps producing foot soldiers and as long as the ideological and social infrastructure that sustains recruitment remains intact, the organizations can persist in a damaged form, can rebuild over time, and can continue to generate attacks. The pipeline arithmetic, in short, closes the door on the cleanest version of the first future and props open the door to the second. It tells us that the shadow war can hurt Pakistan’s terror infrastructure badly and indefinitely, and that it probably cannot destroy it, and a campaign that can wound without killing is a campaign structurally biased toward permanence. That bias is the arithmetic foundation beneath this article’s judgment that the second future, the campaign without end, is the most probable of the four.

The arithmetic deserves one further refinement, because the pipeline is not only a matter of numbers but of the social architecture that produces those numbers, and that architecture is more durable than any count of camps or seminaries can capture. The recruitment system associated with the major anti-India militant organizations is embedded in specific geographies, in particular districts of Pakistani Punjab where decades of activity have produced not merely facilities but a culture, a set of family traditions, mosque networks, charitable fronts, and local prestige hierarchies in which a militant career carries social meaning. Operation Sindoor’s strikes on the Bahawalpur complex and the broader infrastructure at Muridke demonstrated that the physical facilities can be damaged, and damaged severely. What missiles cannot strike is the social process itself, the intergenerational transmission of grievance and purpose that produces a willing recruit before that recruit ever walks into a training camp. A program of targeted killing operates even further from that social process than a missile does. It removes individuals. It does not, and cannot, touch the conditions that produce individuals, and as long as those conditions persist, the production line persists with them.

This is why the leadership-regeneration question is the only place where the arithmetic offers India any genuine leverage, and even there the leverage is limited and contested. Cronin’s assessment of the decapitation pathway is deliberately cautious, and the caution is grounded in a mixed historical record. Some organizations have collapsed after losing key leaders, particularly organizations that were highly centralized, that depended on a charismatic founder, and that lacked a developed second tier. Other organizations have absorbed the loss of senior figures and regenerated, particularly those with deep benches, institutionalized succession, and, crucially, state sponsorship that can shelter and groom replacements. The anti-India militant organizations fall closer to the second category than the first. They are not one-man movements. They have institutional structures, decades of accumulated organizational depth, and, in India’s consistent analysis, the protection and patronage of elements of the Pakistani state, which is precisely the resource that makes leadership regeneration possible. A privately sponsored terrorist group that loses its founders may simply end. A state-sheltered organization that loses its founders can be rebuilt by its sponsor, slowly and at cost, but rebuilt.

The honest synthesis is therefore double-edged. The shadow war’s concentration on senior leadership is the strategically intelligent response to the pipeline arithmetic, the one move available that the recruitment system cannot quickly counter, and sustained pressure on the leadership layer can impose real, lasting degradation on the operational sophistication and cohesion of the targeted groups. But state sponsorship caps the achievable effect. As long as the Pakistani establishment retains both the will and the capacity to shelter and reconstitute militant leadership, decapitation degrades without destroying, and the organizations persist in a wounded but living form. The pipeline arithmetic, fully worked through, delivers a verdict that should discipline every scenario in this article. The shadow war can make Pakistan’s terror infrastructure weaker, less capable, and more frightened, and it cannot, by attrition alone, make that infrastructure cease to exist. A program that can wound indefinitely without delivering a decisive blow is a program whose most probable destiny is to continue indefinitely, and that is the destiny the second future names.

The Next Link

This analysis has been, throughout, an exercise in looking forward, an attempt to read the futures latent in the present. But a forward projection is only half of an understanding, because the futures the shadow war can produce are constrained by the past that produced the shadow war, and the campaign cannot be fully understood as a thing with possible endings until it is also understood as a thing with a definite beginning. The next link in this chain, therefore, is not another event further down the timeline. It is the timeline itself, taken whole.

The shadow war did not begin in March 2022 with the killing of Zahoor Mistry, however convenient that date is as an operational marker. It began on a tarmac in Kandahar in December 1999, when India, to save the passengers of a hijacked airliner, released three militants from its custody, and one of those militants, Masood Azhar, walked free to found Jaish-e-Mohammed. Every killing the shadow war has carried out, every scenario this article has examined, every question about exits and pipelines and escalation, descends from that single decision and from the twenty-six years of attack and response that followed it. The Parliament attack of 2001, the Mumbai siege of 2008, the Pathankot raid of 2016, the Pulwama bombing of 2019, the Balakot airstrike that answered it, the Pahalgam massacre of 2025, and Operation Sindoor: these are not separate stories. They are one story, and the shadow war is one chapter of it, and the futures predicted here are the still-unwritten chapters of the same continuous narrative.

To predict where the campaign goes, then, is to make an implicit claim about the meaning of the whole arc, and that claim deserves to be examined in full rather than smuggled in as an assumption. Is the twenty-six-year sequence from IC-814 to Pahalgam a story of strategic maturation, of a state slowly and painfully learning to defend itself and acquiring, at last, the capability and the will to impose costs on those who attack it? Or is it a story of an escalation spiral, in which each response made the next provocation more likely and the next crisis more dangerous, and in which the shadow war is simply the current rung of a ladder that leads somewhere neither side wants to go? The reader who has followed this article’s four futures will recognize that the answer to that question shapes which future seems most likely, because a maturation reading points toward eventual stability and an escalation reading points toward the fourth future and the open war beyond it.

The maturation reading and the escalation reading are not merely two academic interpretations to be set side by side and left there. They are two different theories of where the shadow war is heading, and a reader who has followed the four futures can now see how each interpretation maps onto them. The maturation reading, in which India has slowly built the capability and will to defend itself, is most compatible with the first and second futures, the futures in which the campaign either achieves a strategic effect or settles into a stable, survivable steady state. The escalation reading, in which each response makes the next crisis more dangerous, is most compatible with the third and fourth futures, the futures of catastrophic exposure and of drift past the boundary of counter-terrorism into interstate covert war. The choice between the two readings is therefore not idle. It is, in effect, a prior belief that shapes which of this article’s futures a given observer will find most credible, and it is precisely because the evidence genuinely supports both readings that the futures cannot be ranked with more confidence than the probabilities offered here allow. An honest forecast inherits the ambiguity of the history it forecasts from.

That question, the meaning of the complete arc, is the subject of the complete chain that this prediction extends and depends upon, the full twenty-six-year narrative told as a single unbroken story from the Kandahar tarmac to the Baisaran Valley. The forward projection offered here is the second-to-last word. The arc itself, traced from its origin, is the last, and it is to that complete telling, and to the foundational account of the shadow war and the survival strategies of the leaders still at large that inform every scenario above, that this chain now turns. What comes next in the shadow war cannot be known with certainty. What can be known is that it will be the next sentence in a story that started long before the first unknown gunman climbed onto a motorcycle in Karachi, and that the story is not finished.

Frequently Asked Questions

How does the shadow war end?

There is no single answer, and any source claiming certainty should be treated with suspicion. The most honest response is that the shadow war can end in one of four broad ways, and that the campaign’s own design makes a clean ending unlikely. It could end if Pakistan verifiably dismantles its terror infrastructure, removing the campaign’s justification, though the structural obstacles to a genuine Pakistani crackdown make this the least probable clean ending. It could end through exposure, if a catastrophic operational failure or the cumulative weight of credible reporting makes continuation too costly. It could escalate past counter-terrorism into something that triggers an open conflict. Or, most probably, it does not end at all in any clean sense, instead persisting indefinitely as a permanent low-intensity feature of the India-Pakistan relationship. The campaign was built without an exit, and a program built without an exit tends not to find one.

Does the campaign have exit criteria?

No, and this is the central structural fact about it. India has never officially acknowledged the shadow war, let alone articulated a condition whose satisfaction would conclude it. The campaign has no stated objective beyond the elimination of wanted individuals on Pakistani soil, and an objective defined as continuous elimination contains no internal point at which it is complete. The closest thing to a criterion is the no-talks doctrine’s demand that Pakistan verifiably dismantle terror infrastructure, but that demand was articulated as a precondition for diplomatic engagement, not as a stand-down condition for covert operations, and verification of a deliberately deniable Pakistani drawdown would be close to impossible in any case.

What are the possible endgame scenarios?

This analysis identifies four. The first is a Pakistani crackdown that removes the campaign’s rationale. The second is indefinite continuation as permanent attrition, judged here the most likely. The third is a catastrophic exposure that forces a halt. The fourth is escalation, in which the campaign succeeds against the militant leadership and then expands its target set toward state officials, changing its character from counter-terrorism into covert interstate warfare. Each carries a different probability and a different set of consequences, and the fourth, while least probable, carries the most severe consequences and therefore demands attention disproportionate to its likelihood.

Which endgame scenario is most likely?

The second, the campaign without end. It is the most probable because it is the only one of the four that requires nothing to happen. The first needs a wrenching Pakistani transformation, the third needs a catastrophic failure, and the fourth needs a deliberate or drifting escalation. The second needs only that the present trajectory continue, and the present trajectory, with rising tempo, climbing target seniority, and a cost-benefit calculation that currently favors continuation, points directly at indefinite persistence. The campaign does not have to be chosen to continue. It only has to not be stopped.

Can permanent attrition succeed against the militant pipeline?

Only partially, and the distinction is crucial. The recruitment and training infrastructure associated with groups like Lashkar-e-Taiba and Jaish-e-Mohammed can replace foot soldiers and mid-tier operatives faster than any covert campaign can eliminate them, because basic indoctrination and training operate on a timescale of months. What the pipeline cannot quickly replace is senior leadership, because experienced commanders and founding figures are the product of careers rather than training cycles. Attrition therefore cannot drain the militant pool from the bottom, but it can degrade the leadership layer at the top faster than that layer regenerates. The campaign can wound the targeted organizations badly and lastingly. It probably cannot destroy them.

What would force India to stop the campaign?

Three things could plausibly force a halt. A verifiable Pakistani dismantlement of terror infrastructure would remove the rationale, though this is unlikely. A catastrophic exposure, a captured operative confessing on camera or a botched strike killing unambiguous innocents, could make the political and diplomatic cost of continuation exceed its strategic value. And sustained pressure from Western partners, several of whom have themselves accused Indian intelligence of extraterritorial operations, could impose a ceiling on how much exposure New Delhi can absorb. Absent one of these, there is no obvious mechanism that forces a stop, which is precisely why indefinite continuation is the most probable future.

Could the campaign escalate to target state officials?

It is structurally possible, though there is no evidence that India has decided to do so, and this analysis does not predict that it will. The concern is not intention but drift. A campaign without exit criteria, having worked through the militant leadership, faces a choice between stopping without any criterion telling it to and continuing by expanding its definition of a target. Because the militant infrastructure is, in India’s own analysis, sponsored and protected by elements of the Pakistani state, the network the campaign maps points toward those state elements. Escalation toward state targets would not require a single dramatic decision. It could occur incrementally, each target marginally more senior than the last, until the cumulative drift has crossed a line no specific decision consciously chose to cross. Killing a state official, unlike killing a designated terrorist, is an act of war by any standard the international system recognizes.

Is permanent shadow war a viable strategic posture between nuclear states?

This is genuinely contested among analysts. The case for viability is that the posture is cheap, domestically popular in India, stable below the conventional threshold, and requires no risky political decisions, and that postures with those properties tend to persist without catastrophe. The case against viability, which scholars including George Perkovich take seriously, is that a posture relying indefinitely on deniability and on staying below thresholds is a posture quietly accumulating escalation risk, and that between nuclear-armed states the accumulation of risk is the accumulation of a catastrophic kind of risk. The honest position is that permanent attrition is stable until it is not, and that its apparent stability to date is evidence that it has not failed yet, not evidence that it is safe.

What would success look like for the shadow war?

This is harder to answer than it sounds, because the campaign has never defined success, and a program that has not defined success has placed success beyond its own reach. If success is understood as the achievement of a state of affairs, the most coherent version would be a Pakistan that has genuinely abandoned the use of militant proxies against India, with the terror infrastructure dismantled and the recruitment pipeline allowed to atrophy. But the shadow war cannot produce that outcome by itself, because attrition cannot dismantle infrastructure or change a sponsoring state’s strategic doctrine. At most the campaign can contribute pressure toward that outcome. Success defined as a destination is therefore not something the campaign can deliver alone. Success defined as continuous activity, the imposition of ongoing costs, is something the campaign delivers easily, but continuous activity is not victory.

How does the no-talks doctrine affect the shadow war’s future?

It removes the campaign’s most obvious off-ramp. In every previous era of India-Pakistan relations, covert pressure could in principle have been wound down as a goodwill gesture before a summit or traded against concessions during a dialogue round, because diplomacy provided a mechanism for converting covert leverage into political settlement. The no-talks doctrine, by refusing engagement until Pakistan verifiably dismantles terror infrastructure, eliminates that mechanism. With no negotiating table, there is no process through which the campaign could be formally retired. The doctrine designed to pressure Pakistan also, as a side effect, traps the shadow war in a state of permanent activity, which is one more reason indefinite continuation is the most probable future.

Is the 2025 ceasefire relevant to the shadow war’s trajectory?

Yes, significantly. The ceasefire that ended the Operation Sindoor exchange in May 2025 halted the conventional confrontation but resolved nothing structural. Pakistan’s terror infrastructure was not dismantled, India’s permanent-confrontation posture was not lifted, and the bilateral relationship remained frozen. The shadow war continued through and after the ceasefire without interruption, which demonstrates that the covert campaign operates on a logic largely independent of the conventional ceasefire’s terms. More importantly, the ceasefire established that India and Pakistan will now exchange conventional fire across the international border under sufficient provocation, which raises the stakes of any future shadow war exposure or escalation, because the conventional response threshold has been demonstrated to be lower than it was before 2025.

Could a change in Indian or Pakistani leadership end the campaign?

It could change the campaign’s tempo and emphasis, but a clean ending is unlikely from a leadership change alone. On the Indian side, the shadow war reflects a broad strategic consensus and substantial public approval for decisive action against cross-border terrorism, so a new government would face strong incentives to continue rather than to visibly stand the campaign down, which could be portrayed as weakness. On the Pakistani side, a new military or civilian leadership willing to genuinely dismantle terror infrastructure could in principle remove the campaign’s rationale, but the structural obstacles to a Pakistani crackdown, the institutional embedding of militant networks and the political cost of appearing to capitulate, would constrain any new leader as much as the current one. Leadership changes shift probabilities at the margin. They do not, by themselves, supply the exit the campaign was built without.

What does catastrophic failure of the campaign look like in practice?

It has several distinct forms. The most damaging would be a captured intermediary who credibly describes, on camera, the chain of payment and tasking that connects an operation to Indian intelligence, converting years of allegation into something close to proof. A second form is a botched strike that kills unambiguous innocents or the wrong target entirely, stripping the campaign of the moral framing that its victims are terrorists. A third is an operation that crosses an unintended line and triggers a response its planners did not anticipate. Any of these would make continuation politically difficult, and in the post-Sindoor environment, where conventional retaliation is a demonstrated option, a humiliating exposure could generate Pakistani domestic pressure for a response that escalates beyond the covert domain entirely.

How has the campaign’s tempo changed over time?

It has risen steadily since inception. The early phase from 2022 was sparse, with widely spaced killings easily dismissed as ordinary criminal or sectarian violence. The year 2023 brought a sharp acceleration, with a dense cluster of killings in the autumn that ended any pretense of coincidence and prompted formal Pakistani accusations. The phase following Operation Sindoor accelerated further still, with 2026 producing eliminations at a rate observers described as unprecedented, more than thirty militants in a single year. A campaign whose operational tempo is rising after four years is not a campaign exhibiting the signs of natural exhaustion or imminent conclusion.

Why can nobody predict the campaign’s end with confidence?

Because the campaign was deliberately built to be unpredictable, and that unpredictability extends to its own designers. A covert program with no public roadmap, no acknowledged objective, and no stated exit criteria offers no anchor for forecasting. The deniability that protects the campaign from attribution also denies analysts the documentary basis for confident prediction. What can be done, and what this analysis attempts, is to map the small number of structurally available futures and to reason about which conditions make each more or less likely. That is forecasting in the only honest sense available, the identification of plausible futures with probabilities attached, not the false precision of a single confident prediction.

Is the shadow war connected to the question of another open India-Pakistan war?

Closely. The covert campaign and the conventional confrontation are not separate categories but phases of one continuous strategic competition. The shadow war’s fourth future, escalation toward state targets, has as its most plausible terminus exactly the open conflict that the covert program was partly intended to make unnecessary. And the broader assessment of whether India and Pakistan will fight again depends in part on whether the shadow war remains contained or drifts toward escalation. The two questions cannot be answered independently, because the same unresolved structural conditions, terror infrastructure intact, bilateral relations frozen, no diplomatic process underway, drive both the persistence of the covert campaign and the risk of the next conventional crisis.

How does the shadow war compare to other states’ targeted-killing programs?

Comparisons are useful but should not be pushed too far. Other states, most prominently the United States with its post-2001 drone campaign and Israel with its long history of targeted operations, have conducted programs that share the shadow war’s logic of eliminating specified individuals on foreign soil. What distinguishes India’s campaign is the combination of total deniability, the reliance on local intermediaries rather than on standoff technology or uniformed personnel, and the complete absence of any public legal or doctrinal framework. The United States constructed an elaborate, if heavily criticized, legal architecture around its program. India has constructed none, because acknowledging the campaign at all would forfeit the deniability on which it depends. This makes the shadow war harder to study, harder to constrain, and harder to predict than its better-documented counterparts, and it is one more reason this analysis can offer structured scenarios rather than confident forecasts.

Will technology change how the shadow war is fought?

Probably, though the change is likely to be gradual rather than sudden. The current campaign is strikingly low-technology at the point of execution, a matter of a recruited gunman and a motorcycle, and that low-technology character is a deliberate choice that maximizes deniability. Over a longer horizon, the tools available to intelligence services, more capable surveillance, biometric tracking, and the broader digitization of how people can be located, will make the identification and tracking of targets easier and may make the program more efficient. But efficiency at the targeting stage does not resolve any of the structural problems this article has examined. Better technology can help the shadow war find and reach its targets. It cannot supply the campaign with an exit, it cannot drain the recruitment pipeline, and it cannot answer the question of what the program is ultimately for. Technology changes the how. It leaves the deeper questions of the why and the when-does-it-end exactly where they are.

Does the shadow war make another conventional conflict more or less likely?

The honest answer is that it does both, depending on which scenario unfolds. In the second future of permanent low-intensity attrition, the campaign arguably reduces the pressure for conventional conflict, because it provides India with a continuous outlet for retaliatory impulse that stays below the threshold of open war, functioning as a release valve. In the fourth future of escalation toward state targets, the campaign does the opposite, becoming the most plausible route to a conventional crisis. The shadow war is therefore not simply a stabilizing or a destabilizing instrument. It is both at once, a release valve in one configuration and a fuse in another, and which role it ends up playing depends on whether the campaign stays contained within its current bounds or drifts past them. That dual nature is one more reason the program resists simple prediction.

What is the single most important thing to understand about the shadow war’s future?

That the campaign has no exit, and that this absence is not an oversight but a structural feature with consequences. A campaign built around a defined objective contains a built-in brake, a point at which achievement signals completion. A campaign built only around continuous elimination contains, instead, a built-in accelerator. It can drift indefinitely, it can escalate incrementally past the boundary of counter-terrorism, and it cannot define its own success. Everything uncertain about the shadow war’s future flows from that one design decision, and any reader who grasps it grasps why the most probable future is not a clean ending of any kind but the quiet conversion of a campaign into a permanent condition.