The Central Intelligence Agency armed the Afghan mujahideen to bleed the Soviet Union, and within a decade the men it had trained, financed, and equipped had become the founding generation of al-Qaeda. Indian intelligence trained and armed the Tamil Tigers to pressure Sri Lanka, and within eight years a Tiger suicide bomber had killed a former Indian prime minister on a campaign stage in Tamil Nadu. These are not obscure footnotes. They are the two most studied cases of what intelligence professionals call blowback, the technical name for the unintended, delayed, and often catastrophic consequences that covert operations inflict on the states that launch them. The pattern they describe is not a possibility that haunts covert action from the outside. It is a structural property built into the activity itself.

That structural property is what makes blowback the single most important analytical lens for understanding the long-term trajectory of India’s shadow war against terror. The campaign of targeted killings that has eliminated dozens of wanted militants across Pakistani cities looks, from the inside, like a clean instrument: a precise, deniable, low-cost method of imposing consequences on men the Pakistani state has sheltered for decades. Every covert program in history has looked clean from the inside at the moment of its launch. The question this analysis sets out to answer is not whether the shadow war will produce unintended consequences. History gives no example of a sustained covert program that did not. The question is what form those consequences will take, when they will arrive, and whether the campaign’s specific design makes it more or less exposed than the programs that came before it.

To answer that question honestly requires setting the shadow war against the historical record rather than against its own self-image. It requires treating blowback not as a moral verdict but as an engineering fact, the way a structural analyst treats metal fatigue. A bridge does not fail because its designers were wicked. It fails because stress accumulates in places the designers did not model, and because the failure arrives long after the design decisions that caused it. Covert operations fail the same way. They accumulate consequences in places their authors did not model, and the bill arrives years after the officials who authorized the program have left office. What follows is an attempt to model the stress points, to name where the shadow war is exposed, and to adjudicate the central disagreement among intelligence scholars about whether blowback can be designed away or whether it is simply the price of the activity.

The Cases: What History Records About Covert Action and Its Aftermath

The word blowback did not begin as a metaphor borrowed from firearms, though the firearms image, gas and debris driven back into the shooter’s face, captures the idea well enough. The term entered the professional vocabulary of intelligence through a classified document. A CIA internal history of the 1953 operation that overthrew Iran’s elected prime minister, Mohammad Mossadegh, used the word to describe the risk that the coup would generate consequences the United States could not anticipate or control. The operation, codenamed TPAJAX, succeeded in its immediate objective. It restored the Shah to unchallenged power and secured Western access to Iranian oil. The blowback the classified history warned about arrived twenty-six years later, when the 1979 revolution swept the Shah away on a tide of anti-American grievance that drew much of its emotional force from the memory of 1953. The hostage crisis at the American embassy in Tehran, the rupture in relations that has lasted to the present, the strategic adversary that Iran became, all of it traces back through a causal chain to a covert operation that was considered, in its own time, a triumph.

The scholar who took the word out of the classified files and built it into a framework was Chalmers Johnson, whose book on the subject, published in the year before the September 11 attacks, gave the concept its modern definition. Johnson defined blowback narrowly and precisely. It was not simply any bad consequence of foreign policy. It was the specific category of consequence that the public could not connect to its cause, because the cause had been secret. When a covert operation produces a delayed effect, the citizens of the sponsoring country experience that effect as an unprovoked attack, a bolt from a clear sky, because they were never told about the operation that provoked it. Blowback, in Johnson’s formulation, is the price a society pays for actions taken in its name but hidden from its knowledge. The secrecy that makes covert action possible is the same secrecy that makes blowback so disorienting when it lands.

Iran’s case repays a closer look, because it establishes a feature of blowback that the headline disasters can obscure. TPAJAX was, by the standards of covert action, a small and economical operation. It did not field an army. It worked through bribery, propaganda, the manipulation of crowds, and the coordination of a handful of Iranian military figures, and it cost a sum that, set against its consequences, was trivial. The operation’s modest scale did not produce modest blowback. The grievance it planted, the conviction among many Iranians that their sovereignty had been violated and their democracy strangled by a foreign hand, did not scale with the operation’s budget. It scaled with the operation’s symbolic meaning. A small, cheap covert action can generate enormous blowback if it touches something a society regards as sacred, and national sovereignty and self-determination are exactly such things. This is a warning the shadow war’s planners should weigh, because the campaign is also a small, economical program by the standards of covert action, and the comfort of small scale is partly an illusion. The blowback a program generates is a function not only of how big it is but of what it touches and how the affected society reads it.

The case that proved Johnson’s thesis with the greatest force is the one that bookends his argument: the American program to arm the Afghan resistance against the Soviet occupation. Operation Cyclone ran from 1979 to roughly 1989, and it grew over that decade into the largest covert action program the CIA had ever undertaken. The numbers are worth stating precisely because their scale is the point. Congressional appropriations and matching Saudi funds channeled an estimated three billion dollars and more in weapons and cash through Pakistan’s Inter-Services Intelligence directorate to the Afghan factions. The program supplied small arms, anti-tank weapons, and, in its final and most consequential phase, the shoulder-fired Stinger anti-aircraft missiles that broke Soviet air superiority over the Afghan valleys. The strategic objective was achieved. The Soviet Union withdrew its last troops across the Amu Darya in February 1989, and the Afghan campaign is widely credited as one of the pressures that hastened the collapse of the Soviet state itself.

That recoil was structural, and it was visible to anyone who cared to look even before the Soviets had finished leaving. The program did not merely supply weapons. It built an infrastructure. It funded the recruitment pipelines that drew tens of thousands of foreign volunteers into the Afghan jihad, and it strengthened the religious schools along the Pakistani frontier that processed those volunteers into fighters. One of the men who organized the foreign-volunteer pipeline, running a recruitment office in Peshawar that funneled Arab fighters toward the Afghan front, was a young Saudi named Osama bin Laden. The organization he built out of that recruitment network, the one he would name al-Qaeda, the base, was a direct institutional descendant of the wartime infrastructure. When the Soviet withdrawal removed the common enemy, the infrastructure did not dissolve. It had its own momentum, its own leadership, its own ideology, and its own arsenal. It turned its attention to new targets. The Afghan civil war of the 1990s, the rise of the Taliban out of the religious schools and refugee camps, the sanctuary Afghanistan provided to al-Qaeda, and finally the attacks of September 11, 2001, all of it grew from soil that the covert program had fertilized. Twelve years separated the last weapons shipment from the towers falling in Manhattan. That gap, the long quiet interval between cause and effect, is the most dangerous feature of blowback, and the analysis will return to it.

The arc of bin Laden’s own life closes the loop of the Afghan case in a way worth pausing on, because it shows how blowback can run for a full generation before it is finally answered. The man recruited and sheltered, in his early career, by the same anti-Soviet effort that the United States bankrolled spent the last decade of his life as the most hunted figure on the planet, and when American forces finally killed him in 2011 they did so in a compound in the Pakistani town of Abbottabad, a short walk from Pakistan’s premier military academy. The story of that decade-long manhunt, told in full in the account of the hunt that ended in Abbottabad, is the closing chapter of a blowback sequence that began with a Cold War decision to arm a resistance. The program that helped create the man required, a generation later, a separate covert operation to remove him. There is no cleaner illustration of blowback’s economics: the original operation looked cheap, three billion dollars to bleed a superpower, and the unbudgeted bill, the wars, the manhunt, the second covert operation, the strategic decade lost, arrived later and dwarfed the original outlay.

India’s own archive contains a case that mirrors the Afghan disaster so closely that it has become the founding trauma of Indian covert doctrine. In the early 1980s, as the ethnic conflict in Sri Lanka intensified, Indian intelligence made a strategic decision to support the Tamil militant groups fighting the Sri Lankan state. The motives were a tangled mixture of genuine sympathy for Sri Lankan Tamils, domestic political pressure from the Tamil Nadu electorate, and a colder strategic calculation about keeping Colombo dependent on New Delhi rather than drifting toward other regional powers. Whatever the precise weighting of those motives, the operational result was concrete. Camps on Indian soil trained Tamil militants. Indian agencies supplied weapons and funds to several Tamil factions, including the group that would eventually devour all its rivals, the Liberation Tigers of Tamil Eelam. For a fuller treatment of how this program was conceived and run, the Sri Lanka operation deserves study as the single most instructive case in the Indian record.

The reversal, when it came, was total. By the Indo-Sri Lankan Accord of 1987, New Delhi had switched sides. The Indian Peace Keeping Force deployed to Sri Lanka to disarm the very militants India had armed, and the Tigers refused to disarm. What followed was three years of brutal urban and jungle warfare in which the Indian military fought an enemy trained, in many cases, by Indian instructors using Indian weapons. The force lost more than twelve hundred soldiers before it withdrew in 1990, its mission unaccomplished and its reputation damaged. The final installment of the blowback arrived on the night of May 21, 1991. At a campaign rally in the town of Sriperumbudur in Tamil Nadu, a young woman approached the former prime minister Rajiv Gandhi, the man who as prime minister had ordered the peacekeeping deployment, bent as if to touch his feet, and detonated an explosive belt. The assassination was a Tiger operation, an act of revenge by the organization India had helped create against the Indian leader it held responsible for turning on it. The full chronology of the movement and its consequences is traced in the Khalistan and separatist violence timeline, which sets the Tamil case alongside other insurgencies and shows how often the pattern repeats. Eight years separated the training camps from the suicide belt. Again, the lethal interval.

The Tamil case deserves its place at the center of Indian covert thinking for a reason beyond its tragedy, and the reason is that the blowback ran in both of the directions the historical record records. It produced a foreign-style blowback, the autonomous proxy turning on its sponsor, the IPKF fighting weapons it had supplied. And it produced a domestic blowback, the assassination of a former head of government on Indian soil, a wound to the country’s own political fabric. The Tamil disaster is, in effect, the Afghan disaster and a piece of the Contra-style domestic damage compressed into a single case and visited upon a single country. That is precisely why it became the founding trauma of Indian covert doctrine rather than a forgotten episode. An institution that has experienced both halves of the blowback pattern in one operation learns the lesson with a thoroughness that no amount of studying other states’ failures could match. The shadow war’s design, as the analysis will later argue, carries the fingerprints of that lesson. The question is whether the lesson learned was the full lesson or only the part most painful to remember.

Two further cases round out the historical record and broaden the pattern beyond the two headline disasters. The American support for the Nicaraguan Contras in the 1980s produced a domestic blowback of a different kind, a constitutional crisis when the covert financing scheme, run partly through illegal arms sales to Iran, was exposed and consumed a presidency in scandal. The lesson there was that blowback is not always foreign and not always violent. Sometimes the consequence is the corrosion of the sponsoring state’s own legal and political institutions, the discovery that running a secret war requires lying to one’s own legislature, and that the lie, once exposed, costs more than the war was worth. The Contra affair is the case that demonstrates blowback’s inward direction. A covert program, to remain covert, must be hidden not only from the adversary but from the sponsoring state’s own oversight institutions, its legislature, its courts, its press. The act of hiding it from those institutions is itself a slow injury to them. When the concealment is later exposed, as concealment on a long timeline almost always is, the injury becomes a public wound, and the public learns that its own government conducted a war it was not told about and was actively misled regarding. The damage is to the trust between a state and its citizens, and that form of blowback is invisible to a purely strategic accounting that counts only foreign consequences.

The Israeli case offers a fourth pattern. In the late 1970s and into the 1980s, Israeli authorities, seeking a religious counterweight to the secular nationalism of the Palestine Liberation Organization, tolerated and in some accounts quietly facilitated the growth of Islamist social and charitable networks in Gaza. The organization that grew out of those networks, Hamas, became by the 1990s a more implacable and more durable adversary than the secular factions it had been encouraged to undercut. The Israeli case is the purest illustration of the proxy paradox: a sponsor builds an actor to weaken an enemy, and the actor becomes a worse enemy than the one it was built to replace. What makes the case especially instructive is that the original calculation was not foolish. A religious social movement did, in the short term, draw energy away from the secular nationalists, exactly as intended. The error was not in the near-term analysis but in the assumption that an actor cultivated for one purpose would remain confined to that purpose. Movements have their own logic. A network built to provide schools and clinics and an alternative to secular politics contained, within that very function, the organizational capacity to become something the cultivators never intended. The proxy paradox is not that sponsors choose badly. It is that the thing chosen has agency, and agency cannot be leased.

Set side by side, these cases share a structure that the rest of this analysis will dissect. Each began as a rational response to a real strategic problem. Each achieved, at least partially, its immediate objective. Each created a capability, an organization, an arsenal, or a grievance that outlived the objective. And in each case the consequence arrived years or decades after the decisions that caused it, by which time the original authors were beyond accountability and the connection between cause and effect had faded from public memory. There is one more shared feature, the most uncomfortable of all. In every case, the officials who launched the program were intelligent, experienced, and acting in what they sincerely believed was the national interest. Blowback is not the signature of incompetence or malice. It is the signature of a structural mismatch between the certainty with which covert decisions are made and the complexity of the systems those decisions disturb. The shadow war is a covert program. It is therefore subject to the same structure. The analytical task is to determine which of these patterns it most resembles, and which of them it has been designed to escape.

The Mechanics: How Blowback Is Born

Blowback is often described as if it were a kind of bad luck, an unforeseeable misfortune that strikes covert programs at random. It is nothing of the kind. The mechanism by which a covert operation generates its own future enemies is regular, observable, and in most cases predictable in outline if not in timing. Understanding that mechanism is the precondition for assessing where the shadow war stands.

The first and most important mechanical fact is that covert operations create durable assets, and durable assets have a will of their own. When a state arms a proxy, it does not simply rent that proxy’s loyalty for the duration of a campaign. It transfers capability. The weapons, once delivered, belong to the recipient. The training, once imparted, cannot be recalled. The organizational skills, the logistics networks, the financing channels, the recruitment pipelines, all of it becomes the permanent property of the proxy, available for whatever purpose the proxy later chooses. The Afghan program is the textbook illustration. The mujahideen factions did not return the Stinger missiles when the Soviets left. The CIA in fact spent years and considerable money in the 1990s running a buyback program, trying to recover the very missiles it had distributed, because a shoulder-fired weapon capable of downing a Soviet helicopter is equally capable of downing a civilian airliner. Once a capability exists in the world, the sponsor cannot un-create it. The proxy becomes an autonomous actor, and an autonomous actor has interests that will eventually diverge from those of its sponsor.

The second mechanical fact concerns the ideological and emotional infrastructure that covert programs build alongside the material kind. To recruit and motivate fighters, a covert program must invest in a narrative. The Afghan jihad was sold to its volunteers as a sacred struggle, a defense of the faith against an atheist invader, and the program’s funders were content to amplify that framing because a religiously motivated fighter is a cheap and committed fighter. The narrative, once propagated, also could not be recalled. The young men who had been told that armed struggle in defense of the faith was the highest of duties did not unlearn that lesson when the Soviets withdrew. They went looking for the next struggle, and the ideological framework the program had helped build pointed them toward new enemies, including the very Western states that had financed their war. A covert program that mobilizes people around an idea is responsible for what that idea does after the program ends.

A third mechanism is the one most specific to the elimination model rather than the proxy model, and it deserves particular attention because the shadow war is an elimination campaign. A targeted killing does not create an armed proxy. It does not transfer a Stinger missile. But it can still generate blowback through a different channel: the channel of grievance, martyrdom, and radicalization. When a militant is shot dead by unknown gunmen, his organization does not simply absorb the loss as a clean subtraction. It frames the death. It builds a narrative around the fallen man, a narrative of martyrdom that can be more mobilizing in death than the man was in life. The history of militant movements is full of figures whose assassination strengthened the cause they served, because a living leader is a fallible administrator while a dead one is a perfect symbol. The shadow war’s planners have to weigh, for every elimination, whether the operational subtraction of one experienced operative is worth the symbolic addition the killing hands to the enemy’s recruiters. The pattern of the targeted killings shows a campaign that has so far concentrated on operational facilitators rather than charismatic figureheads, which suggests the planners are aware of this calculation, but awareness is not immunity.

The fourth mechanism is the one that gives blowback its peculiar power to disorient, and it is a direct consequence of secrecy itself. A covert operation, by definition, is not acknowledged. The sponsoring state denies it. That denial is operationally necessary, but it has a hidden cost. It means the sponsoring state cannot publicly manage, calibrate, or disown the operation. It cannot send a clear signal about where the limits lie, because to define the limits would be to admit the program exists. It cannot reassure a worried ally, or warn an adversary about the consequences of escalation, or correct a misperception, because every such communication would breach the deniability the program depends on. Deniability, in other words, is a one-way door. It protects the program from attribution, but it also strips the sponsor of the tools of public statecraft. When blowback begins to accumulate, the deniable program cannot be steered. It can only be continued or abandoned, and abandoning it is itself an admission. The shadow war is a maximally deniable program. No Indian official has acknowledged it. That deniability is an operational asset, and it is simultaneously a strategic trap, because it means the campaign cannot be publicly calibrated as conditions change.

These four mechanisms, the durability of transferred capability, the persistence of mobilizing narratives, the martyrdom dividend of assassination, and the steering paralysis imposed by deniability, are the engine room of blowback. They are present, in different combinations and intensities, in every case the historical record contains. Not all of them apply equally to the shadow war. The campaign transfers no capability to its enemies, so the first mechanism, the gravest in the Afghan and Tamil cases, is largely absent. But the third and fourth mechanisms apply with full force, and the analysis will show that the campaign’s specific exposure runs through exactly those channels.

There is a fifth mechanism that operates more quietly than the other four, and it is the normalization of method. Every covert operation a state conducts does more than achieve its immediate objective. It also establishes, inside the conducting state’s own institutions, that the method is acceptable, that the threshold for its use has been crossed, that what was once an extraordinary measure is now a tool in the standing repertoire. Each elimination makes the next one easier to authorize, not because anyone decides to lower the bar but because the bar lowers itself through use. An institution that has conducted dozens of targeted killings is a different institution from the one that conducted its first, and the difference is not in capability alone but in the internal sense of what is normal. This matters for blowback because it means a covert program can expand its own scope without any single decision to expand it. The shadow war began, on the available evidence, as a narrowly targeted response to specific militants. The normalization mechanism is the reason such programs rarely stay narrow. Each success widens the implicit definition of an acceptable target, and a widening target definition is itself a form of slow blowback, a consequence the program inflicts on its own judgment.

The mechanisms also interact, and the interactions are where the real danger lies. Consider how the martyrdom dividend and the steering paralysis combine. If an elimination produces an unexpected surge of martyrdom-driven recruitment, a state running an acknowledged program could respond publicly, could recalibrate, could pair the operation with a counter-messaging effort or a visible offer of an alternative path. A state running a deniable program cannot do any of that without breaching the deniability. It is forced to watch the martyrdom dividend accumulate and to respond only with more of the same covert method, because the covert method is the only instrument the deniability leaves available. The mechanisms are not a checklist of separate risks. They are a system, and in a system the failure of one part loads stress onto the others. The shadow war has been engineered to neutralize the first mechanism. It has not been, and by the nature of covert action cannot be, engineered to neutralize the interaction between the mechanisms that remain.

The Delay: Why Consequences Arrive Years Later

If blowback arrived promptly, it would be a manageable problem. A covert program that produced its bad consequences within weeks or months of its bad decisions would generate a clear feedback loop, and clear feedback loops correct themselves. The officials who authorized the program would see the result, learn the lesson, and adjust. The reason blowback is so destructive is precisely that it does not work this way. The interval between the covert action and its consequence is typically measured in years, often in decades, and that delay severs the feedback loop at its most important joint.

Consider the timelines again, laid against one another. The American coup in Iran was executed in 1953; the revolution that constituted its blowback came in 1979, a gap of twenty-six years. The Afghan arms program ran through the 1980s; the attacks that represented its most spectacular blowback came in 2001, twelve years after the program wound down. The Indian support for the Tamil militants peaked in the mid-1980s; the assassination of Rajiv Gandhi came in 1991, between five and eight years after the training camps were most active, depending on which phase of the program one measures from. In no case did the consequence arrive while the decision-makers who caused it were still in a position to learn from it. The official who signs off on a covert program is, on the historical evidence, almost never the official who pays its bill.

This delay has a structural consequence that goes beyond the personal fate of individual decision-makers. It distorts the entire incentive landscape within which covert action is authorized. The benefits of a covert program, its successes, its eliminated targets, its degraded enemy networks, are immediate, visible, and attributable to the officials who ran it. The costs, the blowback, are deferred, diffuse, and by the time they arrive, hard to attribute to any particular decision. An intelligence chief who launches a successful elimination campaign can point to results within the year and be rewarded for them. The blowback from that same campaign may not surface until that chief has retired, written a memoir, and been honored at conferences. The accounting is asymmetric. Visible near-term gains accrue to the decision-maker; invisible long-term costs accrue to the institution and the nation. Any activity with that incentive structure will be systematically over-produced, because the people deciding how much of it to do are insulated from most of its price.

The delay also corrodes public understanding, which is the specific point Chalmers Johnson built his framework around. By the time blowback lands, the covert operation that caused it has usually faded from memory, if it was ever public at all. The citizens who experience the consequence, the embassy seizure, the airliner attack, the assassination, experience it as something inexplicable, an act of motiveless malice that proves the enemy’s barbarism. They are not lying when they say they do not understand why they have been attacked. They genuinely do not, because the operation that supplies the motive was conducted in their name but withheld from their knowledge. This is why blowback so reliably produces escalation rather than reflection. A society that cannot see the causal chain leading to an attack will conclude that the attack was unprovoked, and a society that believes it was attacked without provocation will demand a forceful response, which generates the next covert operation, which plants the seed of the next blowback. The cycle is self-feeding, and the delay is what hides the feeding mechanism from view.

For the shadow war, the delay problem carries a specific and uncomfortable implication. The campaign’s results are being tallied now, in real time. Each elimination is noted, the target’s record recited, the operational success implicitly credited. The campaign therefore looks, by the only measure currently available, like a success. But the only measure currently available is the near-term measure, and the near-term measure is exactly the one that a delayed consequence is structured to evade. If the shadow war is going to generate blowback, the historical pattern says it will not do so on a schedule that allows the present generation of planners to see it. It will surface later, in a form not yet visible, attributable to decisions that by then will be hard to reconstruct. The campaign’s current appearance of clean success is not evidence that it carries no deferred cost. Every program that later recoiled also looked like a clean success at the equivalent stage. The Afghan program looked like the CIA’s masterpiece in 1989. The judgment changed in 2001. The shadow war is, in blowback terms, somewhere around its own 1989. The 2001 has not yet arrived, and the nature of the delay is that no one can be sure it will not.

The delay also creates a specific trap for analysts and commentators, and it is worth naming because the trap is being sprung continuously around the shadow war. The trap is to treat the current absence of catastrophic blowback as a finding rather than as a phase. When years pass and no disaster materializes, the natural inference is that the program is safe, that the feared consequences were never real, that the warnings were the anxiety of people who do not understand operational realities. The historical record exposes that inference as false comfort. The years between the Afghan program and September 11 were not years in which the program was proven safe. They were the incubation interval. The same is true of the years between the Tamil camps and Sriperumbudur. An observer in 1989 who concluded that the Afghan program had produced no blowback would not have been making a cautious empirical judgment; that observer would have been mistaking the quiet before the consequence for the absence of the consequence. The shadow war is now generating exactly this kind of false-comfort commentary, and the delay is the reason the false comfort sounds so reasonable. The correct stance during the incubation interval is not reassurance and not alarm. It is the recognition that the interval is, by definition, the period during which the evidence is not yet in.

One further feature of the delay deserves attention, because it bears directly on how a responsible state should govern a covert program. The delay means that the institution best positioned to detect early blowback is not the institution running the program. The agency conducting the operations is measuring its own success by operational metrics, eliminated targets, intact deniability, and those metrics will look healthy throughout the incubation interval. The early signs of blowback, a shift in recruitment patterns, a hardening of an adversary’s posture, a change in how third countries perceive the sponsoring state, show up in domains the operating agency does not monitor and has no incentive to highlight. This is the structural argument for external oversight of covert programs, oversight by bodies that measure different things and answer to different constituencies. A program watched only by the people running it will receive a clean bill of health for exactly as long as the delay lasts, and then no longer.

The Five Scenarios Facing India’s Shadow War

General observations about blowback are necessary but not sufficient. The useful analytical product is specific: a structured assessment of the particular ways the shadow war could generate unintended consequences, each scenario named, each evaluated for how likely it is and how severe it would be if it occurred. What follows is that assessment. It identifies five distinct blowback pathways. They are not ranked predictions, and they are not mutually exclusive; several could unfold together, and one could trigger another. They are the five stress points where the historical mechanics, applied to the specific facts of this campaign, indicate the structure is most likely to fail.

The first scenario is symmetrical retaliation: Pakistan responding to the shadow war by mounting its own campaign of targeted killings against individuals on Indian soil. This is the most direct form of blowback, and its logic is simple. A method, once demonstrated to work, becomes available to both sides. The shadow war has established, in full view of Pakistan’s security establishment, that motorcycle-borne assassins can reach protected individuals in a hostile country and escape without attribution. Pakistan’s intelligence services possess the capability to attempt the same, and they possess a long roster of individuals inside India, defectors, dissidents, and others, whom they might regard as legitimate targets under a tit-for-tat logic. The probability of this scenario is moderate and arguably rising, because the deterrent against it weakens every time the shadow war demonstrates the method’s effectiveness. The consequence, if it occurred, would be severe: a reciprocal assassination campaign on Indian territory would represent a direct security failure, would be far harder for the Indian state to keep deniable on its own soil, and could escalate toward the kind of open confrontation the covert method was supposed to avoid. The shadow war’s logic assumes India can play this game and Pakistan cannot. That assumption is the scenario’s central vulnerability.

Retaliation also carries a second-order danger that deserves to be drawn out, because it concerns escalation control rather than the killings themselves. A covert exchange between two states, each conducting deniable assassinations on the other’s soil, is a contest with no natural ceiling and no clear communication channel. In a conventional military confrontation, both sides can read each other’s signals, declare red lines, and negotiate de-escalation through established channels. A war fought through deniable assassins has none of that architecture. Neither side acknowledges its operations, so neither side can credibly promise restraint, and an unattributed killing can always be read as either a deliberate escalation or an uncontrolled act by a rogue element. The danger is that a covert exchange, lacking the brakes that conventional confrontations possess, drifts upward through misreading rather than intention. Between two nuclear-armed states, an escalatory drift of that kind is not a minor diplomatic cost. It is the precise scenario that the architects of nuclear stability spend their careers trying to foreclose. The shadow war, by demonstrating and normalizing the deniable-assassination method, lowers the threshold for an exchange whose escalation dynamics are genuinely frightening, and that lowered threshold is a form of blowback even if the retaliation never actually comes.

The second scenario is generational radicalization: the campaign, over time, producing more militancy than it eliminates by handing the enemy’s recruiters a steady supply of martyrs. Every targeted killing removes one operative and creates one funeral. The funeral is a recruitment event. The dead man becomes a name invoked in sermons, a face on a poster, a story told to young men about sacrifice and the implacable enemy who must be resisted. If the campaign continues for years, the cumulative martyrology could become a mobilizing resource for the next generation of militant organizations. The probability of this scenario is genuinely uncertain, and the uncertainty is itself important. It depends on factors the campaign cannot fully control: the resonance of the martyrdom narrative, the supply of disaffected young men, the health of the recruitment networks. The consequence, if the scenario materialized, would be the deepest form of blowback, because it would mean the campaign was net-negative at the level of its own purpose, removing experienced operatives while manufacturing their replacements. This is the scenario the campaign’s target selection appears designed to minimize, by concentrating on operational facilitators rather than figureheads, but minimization is not elimination, and the cumulative effect of many small funerals is hard to forecast.

Radicalization carries a generational structure that makes it especially difficult to detect early. The young man who is drawn toward militancy by the killing of a figure he regarded as a hero does not become a threat immediately. He becomes a threat after years of indoctrination, training, and ascent through an organization. The recruitment effect of an elimination conducted today would therefore not show up in any measurable form until the recruits it produced had matured into operatives, by which time a decade might have passed and the connection between the original elimination and the new generation of militants would be analytically invisible. This is the delay problem from the previous section, applied to the most dangerous of the five scenarios. It means that even if the shadow war is, at this moment, quietly manufacturing its own future enemies, there is no metric currently available that would reveal it. The campaign could be net-negative against its own purpose right now and still appear, by every near-term measure, to be succeeding. The honest position is not that radicalization is certainly occurring. It is that the campaign’s apparent success provides no evidence that it is not.

The third scenario is operational failure resulting in civilian deaths. No covert campaign sustained over years executes every operation cleanly. Eventually an operation goes wrong. A target is misidentified. A bystander is hit. A bomb intended for one man kills a family. The history of targeted killing programs, including the US drone program in Pakistan, is in large part a history of exactly these failures, and the drone program’s reputational damage came overwhelmingly from its civilian casualties rather than from its intended kills. The probability that the shadow war eventually produces a botched operation with civilian deaths is, over a long enough horizon, high; the probability is essentially a function of how long the campaign runs. The consequence would be a sharp shift in the campaign’s international narrative. As long as the targets are demonstrably wanted militants, the campaign enjoys a degree of tacit international tolerance. The first dead child changes that narrative instantly and probably permanently, converting a campaign that could be framed as counter-terrorism into one that critics will frame as state murder. The deniability that protects the campaign also prevents the kind of transparent investigation and accountability that might contain the reputational damage when such a failure occurs.

Botched operations interact with the radicalization scenario in a way that compounds both. A clean elimination of a wanted militant produces a contained funeral, mourned mainly within the militant’s own organization and circle. A botched operation that kills a militant’s children, or kills the wrong family entirely, produces a different kind of grievance, one that radiates outward into communities that had no prior stake in the conflict. The drone program’s experience is again the cautionary text: its civilian casualties did not merely damage its reputation abroad, they generated, in the affected regions, exactly the local hostility that the program’s critics had warned of, turning neutral populations into sympathizers of the militants and, in some documented cases, into recruits. A botched shadow war operation would carry the same risk. It would convert a precise instrument into a blunt one in a single incident, and because the campaign is deniable, the Indian state would have no credible way to express regret, offer compensation, or distance itself from the error. Deniability, which protects the campaign from attribution of its successes, would in that moment also prevent it from disowning its worst failure.

The fourth scenario is the expansion of the diplomatic crisis that the campaign has already begun to generate in the West, and this scenario is distinctive because, unlike the others, it is not hypothetical. It has already started. The killing of a Sikh separatist leader in British Columbia in June 2023, and the subsequent allegation by Canada’s then prime minister in the country’s parliament that Indian government agents had been involved, opened a diplomatic rupture between two democracies that had regarded each other as partners. The crisis deepened when the United States Department of Justice charged an Indian national in connection with an alleged plot to kill a separatist leader on American soil, and deepened again when Canadian police publicly accused Indian diplomats of involvement in a broader campaign of violence, leading to reciprocal expulsions. The extension of the campaign onto Western soil is the clearest evidence available that the shadow war is already producing blowback in real time, and that this particular form of blowback is not delayed at all. It is happening now. The probability of further expansion is therefore not a forecast but an observation of an ongoing process; the relevant question is how far it spreads. The consequence is a category of cost the campaign’s planners may not have priced. Operations against targets in Pakistan carry manageable diplomatic costs because the bilateral relationship is already adversarial and the international community is broadly unsympathetic to Pakistan’s safe-haven record. Operations, or even alleged operations, in Five Eyes democracies carry a different and heavier cost, because they invite the framing of transnational repression, a framing that aligns India rhetorically with the very category of state behavior, the extraterritorial silencing of dissidents, that it would prefer to see attributed only to authoritarian rivals. The damage here is not to a target list but to a national reputation built over decades.

The fifth scenario is exposure: the capture of an operative, the defection of a participant, or the assembly of a forensic and financial trail detailed enough to convert circumstantial attribution into proof. Covert programs are protected by deniability, and deniability is a function of evidentiary ambiguity. That ambiguity is fragile. It survives only as long as no single piece of the operation, no operative, no document, no money transfer, no intercepted communication, can be captured and traced. The longer a campaign runs and the more operations it mounts, the more chances it gives an adversary’s counterintelligence to seize one of those pieces. The American indictment in the alleged plot against the separatist leader on US soil is an early warning of this scenario: it demonstrated that a Western law enforcement system, working a single case, could assemble a detailed narrative naming an alleged Indian official. The probability of a more comprehensive exposure rises with every additional operation, simply as a matter of accumulating exposure. The consequence would be the collapse of the deniability the entire campaign depends on. A proven, documented, attributed targeted killing program is a fundamentally different object, diplomatically and legally, from a suspected one. It would convert every past operation, retroactively, into an admitted act of state, and it would strip future operations of the political cover that deniability provides. The campaign’s greatest asset is the gap between what is suspected and what is proven. The fifth scenario is the closing of that gap.

These five scenarios do not exhaust the possibilities, but they map the principal stress points. Two of them, retaliation and exposure, threaten the campaign’s operational and deniability architecture. One, radicalization, threatens its strategic logic at the root. One, the botched operation, threatens its narrative legitimacy. And one, the diplomatic expansion, is not a threat at all in the conditional sense but a process already underway. An honest assessment of the shadow war has to hold all five in view at once, because the campaign’s apparent success is being measured only against the absence of the first three, while the fourth is already imposing costs and the fifth is accumulating probability with every passing operation.

It is worth noticing what these scenarios have in common, because the common thread is itself an analytical finding. None of the five depends on the campaign failing at its own task. The shadow war could eliminate every man on its target list, cleanly and on schedule, and still suffer all five forms of blowback. Retaliation follows from the campaign succeeding visibly enough to teach the method to the other side. Radicalization follows from the campaign producing the funerals that successful eliminations by definition produce. The botched operation is a statistical property of running many operations, and running many operations is what a successful campaign does. The diplomatic rupture followed from the campaign extending its reach, which is a sign of operational confidence, not operational failure. Exposure becomes more likely the more operations succeed, because each success is another opportunity for evidence to be seized. This is the deepest and most counterintuitive point in the entire blowback framework: a covert program’s blowback is generated by its successes, not by its failures. The shadow war does not have to go wrong to produce its unintended consequences. It only has to keep going right, for long enough, at sufficient scale. That is why the management question cannot be reduced to operational excellence. A campaign can be operationally excellent and still, through that very excellence, be writing the bill that a later decade will be handed.

Inevitability or Design: The Central Disagreement

Among scholars who study covert action, the deepest disagreement about blowback is not whether it happens. The historical record settles that. The disagreement is about whether blowback is inevitable, an unavoidable tax on covert action as such, or whether it is contingent, a consequence of specific design failures that disciplined operations can avoid. The two positions lead to opposite conclusions about the shadow war, and adjudicating between them is the analytical heart of the matter.

One school, the inevitability thesis, holds that blowback is intrinsic to the activity. Its argument runs as follows. Covert action, by its nature, intervenes in complex social and political systems whose behavior cannot be fully modeled. It does so with imperfect information, under time pressure, in pursuit of objectives that are usually narrower than the systems it disturbs. It then conceals what it has done, which prevents both correction and learning. Given those conditions, second-order effects that the planners did not anticipate are not a risk; they are a certainty. The only genuine uncertainty is their form and timing. On this view, the question asked at the start of this analysis, not whether the shadow war will produce blowback but what form it will take, is not rhetorical framing but a literal description of the epistemic situation. The thinkers in this camp would say that the shadow war’s planners can choose the channel through which blowback arrives, by their choice of method and target, but they cannot choose whether it arrives at all.

The design thesis holds the opposite. It argues that the catastrophic blowback cases share specific, identifiable, and avoidable features, and that an operation engineered to lack those features can escape the catastrophic outcome. The avoidable features, on this account, are three. First, the catastrophic cases armed proxies, transferring durable autonomous capability to actors whose future loyalty could not be guaranteed. Second, they pursued open-ended objectives without defined endpoints, so the program had no natural moment of completion and simply accumulated consequences indefinitely. Third, they operated at enormous scale, mobilizing tens of thousands of fighters and billions of dollars, so the infrastructure they built was too large to dismantle. An operation that arms no proxies, that defines limited objectives with clear endpoints, and that operates at small scale, the design thesis argues, removes the mechanisms that produced the disasters. Blowback on this view is not a tax on covert action as such. It is a penalty for a particular, identifiable kind of badly designed covert action.

The shadow war is an unusually clean test of this disagreement, because it was, in part, deliberately designed in reaction to one of the catastrophes. The eliminate-don’t-arm doctrine that the campaign embodies is, on the evidence of India’s evolving covert operations doctrine, a direct institutional lesson drawn from the Tamil disaster. Indian intelligence learned, at the cost of twelve hundred soldiers and a former prime minister, that arming a proxy creates an autonomous actor that can turn. The shadow war’s method, direct elimination by assets that remain under state control rather than the arming of a militant proxy, is precisely the design that the design thesis would prescribe. The campaign transfers no Stinger missiles. It builds no recruitment pipeline for the enemy. It creates no organization that could outlive its purpose and seek new targets. Measured against the specific mechanism that destroyed the Afghan and Tamil programs, the shadow war is genuinely better designed. The first of the four mechanics described earlier in this analysis, the durability of transferred capability, simply does not apply to it. This is a real point in the design thesis’s favor, and it should not be minimized.

There is a subtler version of the design argument that deserves a hearing before the adjudication closes. A defender of the design thesis might concede that the four remaining mechanisms cannot be designed away entirely, but argue that they can be designed down, reduced to a level at which the recoil they generate is small enough to be absorbed as an ordinary cost of statecraft rather than suffered as a catastrophe. On this view, the relevant comparison is not between blowback and its total absence but between a manageable consequence and an unmanageable one, and good design is the discipline that keeps a program on the manageable side of that line. This is a serious argument, and it is largely correct as far as it goes. Tight target selection does reduce the martyrdom dividend. Small scale does limit the footprint. Operational discipline does lower, though it cannot eliminate, the rate of botched operations. The design thesis, in this modest form, is not refuted by the inevitability of recoil; it is simply relocated, from the claim that the consequence can be avoided to the claim that its magnitude can be governed.

But the adjudication cannot stop there, because the design thesis, taken to its conclusion, proves too much. It is true that the shadow war avoids the proxy mechanism. It is not true that avoiding the proxy mechanism means avoiding blowback. The five scenarios set out above are the demonstration. Not one of those five depends on the proxy mechanism. Symmetrical retaliation does not require India to have armed anyone; it requires only that India have demonstrated a method. Generational radicalization does not require a transferred capability; it requires only funerals, and the campaign produces funerals by design. The botched operation requires only the statistical certainty of error over a long campaign. The diplomatic crisis requires only that the campaign cross borders into states that will not tolerate it, and that crisis is already underway. The exposure scenario requires only the slow accumulation of evidentiary opportunity. The shadow war has been engineered against the last war’s blowback, the proxy disaster, with real success. It has not been, and arguably cannot be, engineered against blowback as such, because the remaining four channels are properties of covert action in general rather than of the proxy model specifically.

There is also a reason to doubt that the modest design thesis fully rescues the campaign, and the reason is the diplomatic scenario. Manageable blowback is blowback the sponsoring state can absorb without strategic damage. The diplomatic rupture with Western democracies is not obviously in that category. A reputation as a responsible democratic power, the kind of reputation that secures technology partnerships, defense cooperation, and diplomatic alignment, is a strategic asset built over decades, and the transnational-repression framing threatens it at the root. If the diplomatic blowback metastasizes from a bilateral crisis with one or two countries into a broad Western reassessment of India as a state that conducts extrajudicial killings on allied soil, that is not a manageable cost. It is a strategic loss of exactly the kind the design thesis claims good design prevents. The modest design thesis works for the radicalization and botched-operation channels, where scale and discipline genuinely govern magnitude. It works far less well for the diplomatic channel, where the magnitude is set not by the campaign’s operational discipline but by how the rest of the world chooses to interpret it.

The honest adjudication, then, is that both theses are partly right, and that their partial truths combine into a conclusion sharper than either alone. The design thesis is correct that the catastrophic, civilization-scale blowback, the al-Qaeda outcome, is not the shadow war’s likely fate, because that specific outcome requires the proxy mechanism and the shadow war has designed the proxy mechanism out. The inevitability thesis is correct that a recoil of some form is not escapable, because the channels that remain, retaliation, radicalization, operational error, diplomatic rupture, and exposure, are intrinsic to sustained covert action and cannot be designed away without abandoning the activity. The shadow war has therefore not escaped blowback. It has changed its profile. It has traded the catastrophic-but-classic proxy outcome for a basket of smaller, more diffuse, more various consequences, one of which, the diplomatic, is already being paid. That is a better trade than the Afghan program made. It is not the same thing as a clean escape, and the campaign’s planners do the nation no service if they mistake the one for the other.

Where the Comparison Breaks Down

A comparative analysis has an obligation to mark its own limits, and the comparison between the shadow war and the historical blowback cases breaks down at several specific points that an honest assessment must concede. The historical cases are instruments for thinking, not templates that the present will obediently follow.

The most important breakdown concerns mechanism, and it has already been established but deserves restating as a limit rather than a finding. The Afghan and Tamil programs were proxy programs. They worked by transferring capability to a third actor. The shadow war is an elimination program. It works by directly removing individuals. These are not variants of the same activity; they are different activities with different physics. Any inference that runs from the Afghan disaster directly to a predicted shadow war disaster of the same kind is invalid, because the causal mechanism that produced the Afghan disaster is absent. The comparison illuminates the general principle, that covert action generates unintended consequences, but it cannot be used to predict the specific consequence, because the specific consequences in the historical cases were generated by a mechanism the shadow war does not employ. Readers and analysts who reach for the phrase the next al-Qaeda when discussing the shadow war are misusing the comparison. The valid lesson is structural, not predictive.

A second breakdown concerns scale, and scale is not a minor detail. Operation Cyclone mobilized tens of thousands of fighters and moved billions of dollars over a decade. The Tamil program armed entire militant formations. The shadow war, by every available estimate, is a campaign of individual operations, each involving a small team, conducted at a tempo measured in dozens of operations over several years. The blowback potential of a program is not independent of its scale. A program large enough to reshape the politics of an entire region, as the Afghan program reshaped the politics of South and Central Asia, has a blowback potential commensurate with that footprint. A campaign that removes individuals one at a time has a smaller footprint and, other things equal, a smaller blowback potential. The comparison to the historical giants risks importing their scale into the assessment of a campaign that does not share it. The shadow war’s blowback, when it comes, is more likely to resemble the diplomatic friction it is already generating than the geopolitical catastrophe the Afghan program produced.

A third breakdown concerns the international environment, which is not the environment of the 1980s. The Cold War provided a permissive backdrop for covert action; superpower competition meant that a great deal was tolerated that would not be tolerated now. The contemporary environment is denser with surveillance, with investigative journalism, with the forensic and financial tracing capabilities of advanced states, and with international legal norms around extraterritorial killing that, while weakly enforced, are more developed than they were. This cuts in two directions, and the comparison cannot resolve which dominates. The denser environment makes the exposure scenario more probable than it would have been in the 1980s, which is bad for the campaign. But the same density also means that blowback, when it comes, is more likely to be diplomatic and reputational, the relatively containable kind, than the kind that festers undetected for a decade before erupting. The historical cases cannot tell us how this trade-off resolves, because they unfolded in an information environment that no longer exists.

One further breakdown is the easiest of all to overlook. The historical comparison is built entirely from covert programs that have already produced their unintended consequences, because those are the cases the record preserves and scholars study. There is an inherent selection bias in the archive. Covert operations that produced no significant recoil, or whose effects were small enough to be absorbed without notice, do not generate books and case studies; they generate silence. It is therefore possible that the universe of covert actions includes a substantial number of low-consequence or no-consequence cases that the comparative method never sees, and that the shadow war could belong to that invisible category. This possibility cannot be dismissed, and it is the strongest available argument for the design thesis. It is also, however, the weakest kind of argument, an appeal to the absence of evidence, and it cannot bear much weight. The honest statement of the limit is this: the comparison is built from a biased sample, the sample of failures, and a campaign should not assume it belongs to the unobserved population of successes simply because membership in that population would be convenient.

A further limit concerns attribution itself, and it is a limit that runs in the campaign’s favor and so is easy for a critical analysis to skip. The shadow war is not a confirmed program. No Indian official has acknowledged it, and the evidence connecting India to the eliminations in Pakistan, while substantial in pattern terms, remains circumstantial. This means that every claim in this analysis about the shadow war’s blowback profile is, strictly, a claim about the blowback profile of a campaign whose existence is inferred rather than established. If the inference is wrong, if the pattern of killings has been misread and there is no coordinated Indian campaign, then the blowback framework applies to a phantom. The analysis has proceeded, as the rest of this series does, on the working assumption that the pattern is real and coordinated, because the pattern evidence is strong. But intellectual honesty requires marking the assumption as an assumption. The comparison to the Afghan and Tamil programs is a comparison between two confirmed, documented, admitted covert operations and one inferred one, and the asymmetry in the evidentiary status of the three cases is a real limit on how firmly the comparison can be pressed.

The deepest limit of all is temporal, and it cannot be overcome by any refinement of method. This analysis is being conducted during the shadow war’s incubation interval, the period after the operations have begun but before any blowback they may generate has had time to surface. A comparative study conducted during the incubation interval is necessarily a study of risk rather than a study of outcome. It can identify the channels through which blowback might arrive and assess their relative probability, which is what the five-scenario framework attempts. It cannot report what actually happened, because what actually happened has not yet finished happening. The Afghan and Tamil cases can be analyzed as completed sequences with known endings. The shadow war cannot, and any analysis that claimed to deliver a final verdict on its blowback would be claiming knowledge that the structure of blowback makes impossible to possess at this stage.

What the Comparison Teaches

Setting the shadow war against the long history of covert action and its consequences is not done to deliver a verdict for or against the campaign. That is a judgment that belongs to the political community in whose name the campaign is conducted, and it depends on values, the weighing of security against risk, of present results against future costs, that an analysis of mechanisms cannot settle. The purpose is narrower and, properly understood, more useful. It is to convert blowback from a vague anxiety into a structured set of expectations, so that the people responsible for the campaign, and the citizens on whose behalf it is run, can think about it clearly.

The first thing the comparison teaches is that blowback is a management problem, not a binary outcome. The careless way of thinking about it treats the matter as a switch: either a covert program produces unintended consequences or it does not, and the hope is to be in the lucky category. The historical record does not support that framing. Every sustained covert program produces consequences its authors did not intend; the variation is in the severity, the form, the timing, and above all in how well the sponsoring state anticipated and contained them. The relevant question for the shadow war is therefore not the false binary, will there be a recoil, but the management question, which forms of consequence are most likely, which can be reduced by changes in conduct, and which must simply be planned for. The five-scenario assessment in this analysis is an attempt to put that management question on a usable footing.

A second lesson concerns exit criteria, and it connects directly to the broader study of why covert operations succeed or fail. The catastrophic blowback cases share a feature beyond the proxy mechanism: they were open-ended. The Afghan program had no defined endpoint; it ran until external events ended it, and the infrastructure it had built by then was too large to dismantle. A covert program without exit criteria cannot stop accumulating consequences, because it cannot stop. The shadow war, as far as can be determined, has no publicly articulated endpoint, no statement of the conditions under which it would be judged complete and wound down. A campaign that runs until events end it, rather than until its objectives are met, is a campaign that has surrendered control of its own duration, and duration is the variable most strongly correlated with cumulative blowback. The single most consequential design improvement available to the shadow war’s planners is not a change of method but the definition of an end.

The third lesson is about the asymmetry of the accounting, and it is the lesson most relevant to the citizens rather than the planners. The benefits of the shadow war are being counted now, in public, in the running tally of eliminated militants. The costs, the blowback, will be counted later, in a currency and on a timeline that the present cannot fully see. The diplomatic costs have already begun to accrue in the rupture with Western democracies. The other costs, the retaliation, the radicalization, the botched operation, the exposure, are deferred, and the deferral is not evidence of their absence. It is the defining property of blowback. A society judging a covert program while it is still running is judging it on incomplete information, with the favorable half of the ledger visible and the unfavorable half not yet posted. The mature response to that situation is not paralysis and not denial. It is humility about the provisional nature of the current verdict, and insistence that the program be designed, from the start, with the deferred costs in view.

A fourth lesson follows from the third and concerns the proper role of oversight. If the costs of a covert program are deferred and the benefits are immediate, and if the agency running the program is structurally incapable of detecting the deferred costs because it measures only the immediate benefits, then the case for external scrutiny is not a matter of constitutional propriety alone. It is a matter of operational competence. A program watched only by its operators will be told it is succeeding throughout the incubation interval, and that reassurance will be sincere and wrong. Oversight bodies that answer to different constituencies and measure different things, a legislature, a judiciary, an inspectorate, are not obstacles to a covert program’s effectiveness. They are the only instruments capable of seeing the half of the ledger the operators cannot. The catastrophic blowback cases were, almost without exception, programs that had escaped or evaded that kind of scrutiny. The lesson is not that covert action should never be conducted. It is that covert action conducted without external oversight is covert action that has disabled its own early-warning system, and a program with no early-warning system will discover its blowback only when the blowback arrives in full.

The fifth and final lesson is the one the comparison teaches most insistently and the one most easily resisted, because it asks the campaign’s supporters to hold two thoughts at once. The first thought is that the shadow war is, genuinely, a better-designed covert program than the Afghan or Tamil disasters, that its architects learned the proxy lesson and built a method that closes the gravest blowback channel. The second thought is that being better-designed than two catastrophes is a low bar, and clearing it does not make a program safe. Both thoughts are true, and the temptation is to collapse them, to let the genuine design achievement of the first thought silence the warning of the second. The comparison’s most important teaching is that this collapse must be resisted. A program can deserve credit for its design and still be carrying unpriced risk, and the intellectual discipline blowback demands is the refusal to let the credit cancel the risk. The Afghan program’s planners were not fools, and the Tamil program’s planners were not fools, and their programs still ended where they ended. The shadow war’s planners may well be more careful than either. The history of blowback simply insists that more careful is not the same as safe, and that the interval in which a program looks safe is precisely the interval in which that distinction matters most.

A final observation belongs alongside those five lessons, less a teaching than a caution about how this kind of analysis is read. The comparative method has a built in bias toward the dramatic case. The Iranian coup, the Afghan program, the Tamil sponsorship are remembered precisely because their consequences were spectacular, and a study that draws on them risks implying that every covert program is one missed variable away from a comparable catastrophe. That implication would be false. The larger and quieter part of the intelligence record consists of operations that produced modest, manageable, and mostly forgotten consequences, programs that were wound down on schedule, that stayed small, that never escaped the control of the people who authorized them. Those operations are absent from the casework here not because they did not happen but because survivorship works in reverse for disasters: the failures are studied and the successes are filed. A reader who takes the catastrophic cases as the baseline will overestimate the danger, and an analyst who only ever cites them will have built an argument that cannot distinguish a well run program from a doomed one. The honest position holds both halves together. The dramatic cases prove that the danger is real and that the mechanism is structural. The quiet cases prove that the danger is not destiny, that conduct and design and restraint genuinely change the odds, and that a program built by people who have studied the failures has a materially better chance than one built by people who have not. The shadow war should be judged against the whole distribution of outcomes, not against its worst tail alone, and the discipline this analysis asks for is the same discipline it has asked for throughout: neither the false comfort of assuming the quiet cases are the rule, nor the false dread of assuming the catastrophes are.

The CIA armed the Afghan mujahideen, and the mujahideen became al-Qaeda. Indian intelligence trained the Tamil Tigers, and the Tigers killed Rajiv Gandhi. These two sentences are the permanent backdrop against which any honest assessment of the shadow war has to be written. They do not prove that the shadow war will end in disaster; the campaign has, genuinely, been designed to avoid the specific mechanism that produced those two disasters. But they do prove that the confidence of a covert program’s planners at the moment of its apparent success is worth nothing as a predictor of its eventual cost, because the Afghan program’s planners were just as confident in 1989. The shadow war may well be the better-designed campaign its architects believe it to be. The history of blowback simply insists that this is a claim to be tested over decades, by consequences not yet visible, and not a verdict that the present is entitled to deliver in the campaign’s favor. The unintended consequence is coming. The discipline the comparison demands is to keep watching for it, to keep the campaign small and bounded enough that the consequence stays containable, and never to mistake the quiet of the present interval for the safety of the whole.

Frequently Asked Questions

Q: What is blowback in covert operations?

Blowback is the term intelligence professionals use for the unintended, delayed, and often damaging consequences that a covert operation inflicts on the state that launched it. The word entered the professional vocabulary through a classified CIA history of the 1953 Iran coup, and the scholar Chalmers Johnson later gave it its modern definition. The crucial element of that definition is secrecy: blowback is not simply any negative result of foreign policy, but specifically the negative result that the public cannot connect to its cause, because the cause was a hidden operation. A society subjected to blowback experiences it as an unprovoked attack, because it was never told about the covert action that provoked it. That disconnect between visible effect and concealed cause is what makes blowback both distinctive and dangerous.

Q: Where did the term blowback originate?

The word was first used in its intelligence sense in a classified internal CIA document, a history of the 1953 operation that overthrew Iran’s elected prime minister, Mohammad Mossadegh. The document used blowback to describe the risk that the coup would generate consequences the United States could neither anticipate nor control. The image is drawn from firearms, the gas and debris that can be driven back toward the shooter, and it captures the idea that a covert action can recoil onto its author. The term remained largely confined to professional circles until the scholar Chalmers Johnson built a public framework around it, publishing a book on the subject in the year before the September 11 attacks, which gave the concept wide currency.

Q: How did the CIA’s Afghan program lead to al-Qaeda?

Operation Cyclone, the CIA program to arm the Afghan resistance against the Soviet occupation, ran through the 1980s and channeled an estimated three billion dollars and more in weapons and funds through Pakistan’s intelligence service to the Afghan factions. The program did not only supply weapons; it helped build an infrastructure of recruitment pipelines and religious schools that drew tens of thousands of foreign volunteers into the conflict. One organizer of the foreign-fighter pipeline was Osama bin Laden, and the organization he built from that wartime network, al-Qaeda, was a direct institutional descendant of it. When the Soviet withdrawal removed the common enemy, the infrastructure did not dissolve. It retained its leadership, ideology, and arsenal, and it turned toward new targets, culminating in the attacks of September 11, 2001, twelve years after the program ended.

Q: Did India’s support for the Tamil Tigers cause Rajiv Gandhi’s assassination?

The causal chain is direct, though it ran through a reversal of policy. In the early to mid-1980s, Indian intelligence trained and armed Tamil militant groups, including the Liberation Tigers of Tamil Eelam, as part of its approach to the Sri Lankan conflict. By the 1987 Indo-Sri Lankan Accord, India had switched sides and deployed a peacekeeping force to disarm those same militants. The Tigers refused, and three years of fighting cost the Indian force more than twelve hundred soldiers. On May 21, 1991, a Tiger suicide bomber killed Rajiv Gandhi, the former prime minister who had ordered the peacekeeping deployment, at a campaign rally in Tamil Nadu. The assassination was an act of revenge by an organization India had helped create against the leader it held responsible for turning on it, which makes it one of the clearest cases of blowback in the historical record.

Q: What are the biggest examples of covert-operation blowback?

The two most studied cases are the American program to arm the Afghan mujahideen, which contributed to the rise of al-Qaeda, and India’s support for Tamil militants, which led to the assassination of Rajiv Gandhi. Two further cases broaden the pattern. The American support for the Nicaraguan Contras produced a domestic constitutional crisis when its illegal financing was exposed, showing that blowback can corrode the sponsoring state’s own institutions rather than producing a foreign attack. And Israel’s tolerance of Islamist networks in Gaza as a counterweight to secular Palestinian nationalism contributed to the rise of Hamas, illustrating the proxy paradox in which a sponsor builds an actor that becomes a worse adversary than the one it was meant to undercut.

Q: Could India’s shadow war produce blowback?

Yes, and the relevant question is not whether but in what form. The shadow war is a sustained covert program, and the historical record contains no example of a sustained covert program that produced no unintended consequences. The campaign has been designed to avoid the specific mechanism that destroyed the Afghan and Tamil programs, the arming of proxies, and that design choice is real and significant. But avoiding the proxy mechanism does not avoid blowback as such, because other channels remain open: symmetrical retaliation, generational radicalization, operational failure, diplomatic rupture, and the eventual exposure of operations. At least one of these, the diplomatic, is already imposing costs through the rupture in relations with Western democracies.

Q: What are the most likely blowback scenarios for India?

A structured assessment identifies five. The first is symmetrical retaliation, with Pakistan mounting its own targeted killings on Indian soil. The second is generational radicalization, with the campaign’s funerals serving as recruitment events for future militancy. The third is operational failure, the statistically near-certain eventual botched operation that kills civilians. The fourth is the expansion of the diplomatic crisis already underway with Western states. The fifth is exposure, the capture of an operative or the assembly of an evidentiary trail detailed enough to convert suspicion into proof. The scenarios are not mutually exclusive, and the fourth is not a forecast but an observation of an ongoing process.

Q: Has the Canada diplomatic crisis been blowback from the shadow war?

It is the clearest available evidence that the shadow war is already producing blowback, and that this particular form arrives with no delay at all. The killing of a Sikh separatist leader in British Columbia in June 2023, followed by a Canadian prime minister’s parliamentary allegation of Indian government involvement, opened a rupture between two democracies. It deepened with a US Department of Justice indictment in an alleged plot on American soil and with Canadian police accusations against Indian diplomats, leading to reciprocal expulsions. This is blowback in Chalmers Johnson’s precise sense, an unintended consequence flowing from covert action, except that unlike the classic delayed cases, this one is unfolding in real time and in full public view.

Q: Could Pakistan retaliate with its own targeted killings in India?

It is one of the five identified scenarios and arguably the most direct. The logic is that a method, once demonstrated to work, becomes available to both sides. The shadow war has shown, in full view of Pakistan’s security establishment, that assassins can reach protected individuals in a hostile country and escape without clear attribution. Pakistan’s intelligence services possess comparable capability and a roster of individuals inside India they might regard as legitimate targets under a tit-for-tat logic. The deterrent against this scenario weakens each time the campaign demonstrates the method’s effectiveness. The shadow war’s strategic logic assumes India can play this game while Pakistan cannot, and that assumption is the scenario’s central vulnerability.

Q: Why does blowback take so long to appear?

The delay is structural and is the feature that makes blowback so destructive. The historical intervals are long: twenty-six years between the Iran coup and the revolution, twelve years between the end of the Afghan program and the September 11 attacks, several years between the Tamil training camps and Rajiv Gandhi’s assassination. The delay severs the feedback loop, because the officials who authorize a covert program are almost never the ones in office when its consequences arrive. It also distorts incentives: the benefits of covert action are immediate and attributable while the costs are deferred and diffuse, so the activity tends to be over-produced. And the delay corrodes public understanding, because by the time blowback lands, the operation that caused it has faded from memory or was never public, so the consequence is experienced as an unprovoked attack.

Q: What is the difference between blowback and ordinary retaliation?

Ordinary retaliation is a direct, acknowledged, and usually prompt response to a known act, and both sides understand the causal chain. Blowback is different on each of those dimensions. It is typically delayed by years or decades, it flows from an action that was covert and therefore unacknowledged, and crucially the public of the sponsoring state cannot connect the consequence to its cause. That disconnect is the defining feature. When a society suffers blowback, it genuinely does not understand why it has been attacked, because the provoking operation was concealed from it. Ordinary retaliation invites a measured response; blowback, because it appears unprovoked, tends to invite escalation, which generates the next covert operation and the next cycle.

Q: Is blowback inevitable or preventable?

Scholars genuinely disagree, and the honest answer combines both positions. The inevitability thesis holds that blowback is intrinsic to covert action, because it intervenes in complex systems with imperfect information and then conceals what it did, preventing correction. The design thesis holds that catastrophic blowback comes from specific avoidable features, arming proxies, pursuing open-ended objectives, and operating at vast scale, and that an operation lacking those features escapes the catastrophic outcome. The adjudication is that both are partly right. The shadow war has designed out the proxy mechanism, so the civilization-scale, al-Qaeda-type outcome is unlikely. But the remaining recoil channels are properties of covert action in general and cannot be designed away. The campaign has changed its risk profile, not escaped blowback.

Q: Can a targeted killing campaign radicalize more people than it removes?

It is a genuine risk and one of the five identified scenarios, though its probability is uncertain. Every elimination removes one operative and creates one funeral, and a funeral can function as a recruitment event. The dead man becomes a martyr, a name in sermons and a face on posters, and over a long campaign the cumulative martyrology can become a mobilizing resource for the next generation of militancy. Whether the campaign is net-negative on this measure depends on factors it cannot fully control, such as the resonance of the martyrdom narrative and the supply of disaffected recruits. The shadow war’s apparent focus on operational facilitators rather than charismatic figureheads suggests its planners are aware of the calculation, but awareness reduces the risk without eliminating it.

Q: What would the worst-case blowback scenario for India look like?

The deepest form of blowback would be the radicalization scenario reaching the point where the campaign manufactures militant capability faster than it removes it, making the entire effort net-negative against its own purpose. A second severe outcome would be the exposure scenario producing comprehensive, documented proof of the program, which would collapse its deniability and convert every past operation retroactively into an admitted act of state. A third would be symmetrical retaliation escalating into the kind of open confrontation the covert method was designed to avoid. None of these is the most probable outcome, and the campaign’s design makes the civilization-scale catastrophe of the Afghan precedent unlikely, but worst-case planning requires naming them honestly rather than assuming the favorable path.

Q: Does deniability make blowback worse?

Deniability is a double-edged instrument. It protects a covert program from attribution, which is its operational purpose, but it also strips the sponsoring state of the tools of public statecraft. A deniable program cannot be publicly calibrated, because to define its limits would be to admit it exists. The sponsor cannot signal restraint to an adversary, reassure a worried ally, or correct a dangerous misperception without breaching the secrecy the program depends on. Deniability is therefore a one-way door: it lets the program operate, but it removes the ability to steer the program as conditions change. When blowback begins to accumulate, the deniable program cannot be adjusted; it can only be continued or abandoned. In that sense deniability does make blowback harder to manage.

Q: How does the eliminate-don’t-arm doctrine reduce blowback?

The gravest blowback mechanism in the historical record is the transfer of durable capability to a proxy. When a state arms a militant group, the weapons, training, and organizational skills become the permanent property of that group, available for whatever purpose it later chooses, and the group becomes an autonomous actor whose interests will eventually diverge from the sponsor’s. The Afghan and Tamil disasters both ran through this mechanism. A campaign built on direct elimination by assets that remain under state control transfers no capability, builds no enemy organization, and creates nothing that can outlive its purpose and seek new targets. The shadow war’s eliminate-don’t-arm design, itself a lesson drawn from the Tamil disaster, genuinely closes that channel. It does not close the other channels through which blowback can still arrive.

Q: How does India manage blowback risk?

The available evidence suggests the principal management tool has been design rather than active mitigation: the choice of an elimination model over a proxy model, which removes the most catastrophic blowback mechanism, and an apparent preference for targeting operational facilitators rather than charismatic figures, which limits the martyrdom dividend. What is harder to identify is evidence of management for the remaining risks. There is no publicly articulated endpoint for the campaign, and exit criteria are the single most important safeguard against cumulative blowback, because a program that cannot stop cannot stop accumulating consequences. Effective blowback management would require defining the conditions under which the campaign would be judged complete, keeping its scale bounded, and treating the deferred costs as real rather than hypothetical.

Q: Has any country avoided blowback from a covert program?

This is the hardest question to answer honestly, because of a selection bias in the historical record. Covert operations that produced significant blowback generate books, investigations, and case studies; operations whose blowback was small enough to absorb generate silence. It is therefore possible that a substantial number of low-blowback covert actions exist that the comparative method never sees, and a covert program might belong to that less-visible category. But this is an argument from the absence of evidence, and it cannot bear much weight. The responsible conclusion is that the historical archive is built from a biased sample of failures, that no covert program should assume membership in the unobserved population of successes simply because it would be convenient, and that the prudent course is to design for blowback rather than to hope for its absence.