A little before eight in the morning on a cold January day in 2012, a green Peugeot 405 inched through the rush-hour crush near Gol Nabi Street in northern Tehran. A motorcycle slid up alongside it, the rider reached out with a gloved hand, and a magnetic device the size of a paperback book attached itself to the driver’s door. Seconds later the rider was gone, threading the gridlock that no police car could follow, and the man inside the Peugeot, a thirty-two-year-old chemical engineer named Mostafa Ahmadi-Roshan, was dead. He had been a deputy director at the Natanz enrichment site. He was the fifth Iranian connected to the nuclear program to die in this exact manner in four years, and the pattern by then was no longer in dispute. Someone had built a machine for killing scientists in traffic, and it was running with industrial reliability.

That machine is the subject of this account. For roughly fifteen years, an unacknowledged campaign hunted the men who built Iran’s capacity to enrich uranium, eliminating them one at a time on the streets of their own capital, then sabotaging the centrifuges they had spun, then, in the summer of 2025, abandoning deniability altogether and bombing their homes from the air. The effort is overwhelmingly attributed to Israel and its foreign intelligence service. It has never been formally claimed in full. And it matters here, on a site devoted to a very different conflict, because it is the closest working model the world has produced for what India has been doing inside Pakistan. The motorcycle, the magnetic charge, the deniability, the targeting of human capability rather than infrastructure, the official silence paired with unofficial pride: every signature element of India’s shadow war against terror appears first, fully formed, in the streets of Tehran.

To read the Iranian effort closely is therefore to read a manual that New Delhi appears to have studied. By the end of this piece you will have a scientist-by-scientist account of who died and how, an explanation of the parallel sabotage track that ran alongside the killings, a sober assessment of whether any of it actually delayed the thing it was meant to delay, and a structured comparison that places each Iranian operation against its closest counterpart in Pakistan. None of this is offered as celebration. The Iranian operation killed men in front of their children, it almost certainly killed at least one person by mistake, and it raised questions about sovereignty and law that remain unresolved. But it is the precedent, and India’s planners did not invent their doctrine in a vacuum. They inherited a template, and the template was tested first against Iran.

Background and Triggers

Iran’s pursuit of nuclear technology did not begin as a secret. It began, improbably, as an American gift. Under the Shah, with Washington’s encouragement, Iran launched a civilian nuclear program in the 1950s as part of the Atoms for Peace initiative, and through the 1970s the monarchy signed contracts with American, French, and German firms to build reactors. The 1979 revolution froze all of it. The new theocratic government initially regarded nuclear power as a decadent monarchical vanity, and the brutal eight-year war with Iraq consumed every available resource. Only in the late 1980s, with Ayatollah Khomeini’s death and the rise of a more pragmatic leadership, did the program revive, and this time it revived with a harder edge.

What changed the calculation was the war itself. Iraq under Saddam Hussein had used chemical weapons against Iranian soldiers and Kurdish civilians, and Western governments had largely looked away. Tehran absorbed a brutal lesson: in a region where adversaries reach for unconventional weapons and the international community shrugs, a state that cannot deter is a state that bleeds. Iran also watched as the world discovered, after the 1991 Gulf War, that Saddam’s clandestine nuclear program had advanced far further than inspectors believed. The Iranian leadership drew its own conclusion from that revelation, and it was not the conclusion the West hoped for. The lesson Tehran took was not that secret programs get caught. It was that secret programs must be buried deeper.

Through the 1990s Iran rebuilt its scientific base with help from an unlikely supplier. The Pakistani metallurgist Abdul Qadeer Khan, architect of his own country’s bomb, ran a private proliferation network that sold centrifuge designs and components to Libya, North Korea, and Iran. The centrifuge blueprints that would later spin at Natanz traced their lineage to Khan’s network, which is one of the quieter ironies of South Asian security history: the same Pakistani scientific establishment that this series examines elsewhere also seeded the Iranian program that Israel would spend two decades trying to dismantle. By 2002 the secret was out. An Iranian opposition group revealed the existence of a large enrichment facility under construction at Natanz and a heavy-water plant at Arak, and the International Atomic Energy Agency began the long, frustrating process of inspection, negotiation, and crisis that has defined the file ever since.

For Israel, the Natanz disclosure was a strategic earthquake. Israeli planners had spent decades treating the prevention of a hostile nuclear neighbor as an existential imperative, a doctrine sometimes called the Begin Doctrine after the prime minister who ordered the 1981 destruction of Iraq’s Osirak reactor. The same logic produced the 2007 airstrike on a Syrian reactor at Al Kibar. But Iran presented a problem that neither Iraq nor Syria had. Iran’s program was not a single building. It was dispersed across the country, much of it buried under mountains, hardened against exactly the kind of airstrike that had worked against Osirak. The enrichment hall at Natanz sat beneath meters of reinforced concrete and earth. A later facility near the holy city of Qom, called Fordow, was tunneled into a mountain so deep that even the largest conventional bombs in the American arsenal might not reach it. Israel looked at this geography and concluded that bombing the program flat, in a single decisive raid, was no longer realistic.

If the buildings could not be destroyed, the planners reasoned, then perhaps the program could be attacked through its other vulnerabilities. A nuclear program is not only concrete and centrifuges. It is also knowledge, and knowledge lives in people. A centrifuge cascade requires metallurgists who understand the alloys, engineers who understand the cascades, physicists who understand the neutron behavior, and managers who understand how to weld an entire enterprise together under sanctions and surveillance. These specialists are not interchangeable. Training a senior centrifuge engineer takes years, and a program that loses several of them in a short span loses not only their hands but their judgment, their institutional memory, and their willingness to keep showing up to work. Israeli intelligence reportedly concluded that the human layer of the Iranian program was its softest target, and that a campaign against individuals could buy something that diplomacy and sanctions had not: time.

This was the strategic trigger. It was also a doctrinal inheritance. Israel had been killing its enemies on foreign soil since the 1950s, and it had a settled institutional confidence about doing so. The predecessor campaign that hunted the Munich Olympics killers across Europe in the 1970s had already established that the Israeli service would pursue named individuals for years, across borders, with patience. The Buenos Aires operation that seized Adolf Eichmann two decades earlier had established that Israeli planners felt no particular deference to the sovereignty of states they considered complicit or hostile. The Iranian scientist campaign did not require Israel to invent a new philosophy. It required only that the existing philosophy be pointed at a new category of target: not terrorists, not war criminals, but engineers and physicists, men with no blood on their hands in any conventional sense, whose offense was competence in service of a program their government insisted was peaceful.

That last point is where the operation becomes genuinely contested, and an honest account has to sit with the discomfort rather than wave it away. The men who died in Tehran traffic were not gunmen. Several had teaching posts. Some had families who insisted, plausibly or not, that their relatives worked only on civilian energy or academic physics. Iran has always maintained that its program is for power generation and medicine and that it has never sought a weapon, and while Western intelligence agencies and the IAEA have documented activities difficult to reconcile with a purely civilian purpose, the legal and moral status of killing a scientist is not the same as the status of killing a uniformed combatant or a designated terrorist. The effort’s defenders argue that a scientist knowingly building a weapon for a state that has threatened to annihilate Israel is a legitimate target of preemptive self-defense. Its critics argue that this logic, accepted broadly, would license the assassination of weapons researchers in every country on earth. That argument is not resolved here, and it is not resolved anywhere, which is itself one of the reasons the precedent travels so easily to other conflicts.

A further point of background deserves emphasis, because it shapes everything that follows. Israel did not pursue the scientists in isolation from the wider diplomatic picture. The killings unfolded against a backdrop of negotiation, sanctions, and recurring crisis at the United Nations and the IAEA. Through the late 2000s and into the 2010s, the major powers alternately threatened Iran and bargained with it, layering on financial penalties while leaving a diplomatic door ajar. The covert track operated underneath this public theater, and it operated precisely because the public theater kept failing. Sanctions hurt the Iranian economy without halting the centrifuges. Negotiation produced agreements that collapsed or were never reached. The assassination effort was, in this sense, the instrument states reach for when the visible tools have not worked and the alternative, open war, is judged too dangerous. That position, between failed diplomacy and unwanted war, is exactly the position India occupies with respect to Pakistan, and it is the structural reason the two stories rhyme. Neither New Delhi nor Tel Aviv chose covert killing as a first preference. Each arrived at it after concluding that the polite instruments had been exhausted and the violent alternative was worse. Understanding that sequence matters, because it explains why a covert program is so hard to abandon once it begins: the moment a state stops, it is left again with only the failed diplomacy and the unwanted war, the two options that drove it to covert action in the first place. A program of this kind is a trap as much as a tool. It is easier to start than to end, and a state that runs one for fifteen years, as Israel did against Iran, will find that the operation has become load-bearing in a strategy that has no obvious replacement for it.

The First Strikes: Hosseinpour and the Quiet Years

The effort did not announce itself. Its opening was ambiguous enough that, even now, analysts disagree about where to date the beginning. The most commonly cited first death is that of Ardeshir Hosseinpour, a specialist in electromagnetism and a figure associated with the uranium conversion facility at Isfahan, who died in January 2007. Iranian authorities initially attributed his death to gas poisoning, a mundane domestic accident. An American private intelligence firm later reported that he had in fact been killed by Israeli operatives, and that the gas-poisoning story was a cover. Iran denied the assassination claim, the firm’s reporting could not be independently confirmed, and Hosseinpour’s death has floated ever since in a zone of uncertainty: probably the first, possibly not connected at all.

That ambiguity is itself instructive, and it is worth pausing on, because it recurs throughout this story and throughout the Pakistani campaign that this series tracks. A well-run covert program does not produce clean, attributable events. It produces deaths that could be accidents, illnesses that could be natural, and car crashes that could be car crashes. The early Iranian deaths were deniable not because anyone tried hard to deny them but because they were genuinely ambiguous, and ambiguity is a form of operational security more durable than any cover story. When Pakistani clerics began dying of sudden cardiac events and unexplained illnesses in the 2020s, Indian planners were applying a lesson that the Hosseinpour case had taught fifteen years earlier: the most secure assassination is the one nobody is certain was an assassination at all.

For roughly three years after Hosseinpour, the operation, if it was a campaign, stayed quiet. Then in January 2010 it became unmistakable. Masoud Alimohammadi, a fifty-year-old professor of elementary particle physics at Tehran University, left his home in the Qeytariyeh district on the morning of January 12 and was killed when a bomb attached to a motorcycle parked outside his house detonated by remote control. The Iranian government immediately blamed Israel and the United States. What complicated the story was Alimohammadi’s profile. He was a quantum field theorist, an academic by every visible measure, and there was no public evidence tying him to the enrichment program. Some analysts argued he had been targeted for a hidden role; others suggested the killing was meant as a message, or that Iranian intelligence itself was involved in a factional dispute, a theory Tehran later seemed to encourage by parading a confession that few outside Iran believed.

The Alimohammadi killing established the effort’s signature method in the form it would keep for years. The weapon was small, placed by hand or by a passing rider, and detonated remotely or by magnetic adhesion to a moving vehicle. The setting was a routine commute, the most predictable window in any target’s day. The escape relied not on speed alone but on the specific chaos of Tehran traffic, where a motorcycle is one of millions, anonymous by sheer density, and where a pursuing vehicle is simply stuck. These were not improvised choices. They were the output of a doctrine that had identified, with some precision, the conditions under which a city becomes an ally of the assassin rather than the target. India’s planners would later identify the same conditions in Karachi and Lahore, and this series has examined the resulting motorcycle methodology in detail. The Tehran operations were the proof of concept.

It is worth being precise about why the motorcycle works, because the reasoning is identical across both campaigns. A motorcycle threads congested traffic that a car cannot. It carries two riders, allowing one to drive while the other places a charge or fires a weapon. It is mechanically trivial to abandon and burn. And in a large South Asian or Middle Eastern city it is visually invisible, because there is no such thing as a suspicious motorcycle when there are millions of them. The Iranian operation demonstrated all four advantages before India ever applied them, and the consistency of the method across continents tells you something important. Either the same doctrine was deliberately transferred, or independent professionals studying the same problem arrived at the same answer. Both explanations point to the same conclusion: the motorcycle assassination is not a cultural quirk. It is the rational solution to a specific tactical problem, and it will keep reappearing wherever that problem exists.

What the quiet years also reveal is the patience a serious covert program demands. Between Hosseinpour in 2007 and Alimohammadi in 2010 there were three years of apparent inaction, and inaction of that length is not the same as failure. A target must be identified, his routine mapped over weeks, his residence and route and habits documented, local assets recruited or positioned, weapons assembled and emplaced, and an escape rehearsed. Each of these steps is slow, and slowness is a feature rather than a bug, because haste is how covert operations get exposed. The three-year gap was almost certainly not a pause but preparation, and the lesson for any state running a similar program is that the visible tempo of killings reveals very little about the invisible tempo of surveillance underneath. India’s modus operandi of covert eliminations, examined elsewhere in this series, shows the same iceberg structure: a small number of visible strikes resting on a large, slow, hidden body of preparatory work. A reader watching only the strikes sees a handful of events spread across years and might conclude that the operation is sporadic or opportunistic. The reality underneath is continuous. Surveillance teams, handlers, safe houses, and logistics chains run without interruption, and the gaps between killings are not idleness but the time it takes to bring the next target from identification to vulnerability. This distinction matters for anyone assessing India’s shadow war, because the same misreading is available there. The Pakistani killings arrive in clusters and lulls, and the lulls invite the assumption that the operation has paused. The Iranian precedent suggests the safer assumption is the opposite: that the quiet stretches are when the most important and least visible work is being done.

The 2010 Wave: Bombs on Tehran’s Streets

If 2010 opened with ambiguity, it closed with a demonstration of reach that left no room for doubt. On the morning of November 29, two attacks unfolded within minutes of each other in different parts of Tehran, and the coordination announced that whoever was running the operation could strike multiple targets in a single window.

In the first attack, Majid Shahriari, a nuclear engineer and a faculty member at Shahid Beheshti University, was driving through northern Tehran when motorcyclists drew alongside his car and attached a magnetic device to it. The charge detonated and killed him. His wife, in the car, survived with injuries. Shahriari was, by Iranian accounts, a figure of real significance to the program, involved in work that touched the enrichment effort directly, and his death removed a genuine node of expertise rather than a symbolic one.

In the second attack, almost simultaneously and a few kilometers away, the same method was used against Fereydoon Abbasi-Davani, a physicist with a long association with the program and, by many assessments, a more central figure than Shahriari. Abbasi survived. Accounts of his survival vary, but the most cited version holds that Abbasi, who had reason to be alert to the danger, recognized the motorcycle’s approach or heard the device attach, and pulled himself and his wife out of the car before the charge fully detonated. His instinct, sharpened by knowing he was a likely target, saved his life. Iran’s response to Abbasi’s survival was telling. Within months, the government appointed him head of the Atomic Energy Organization of Iran, a deliberate and public promotion of a man the operation had tried and failed to kill. The message was that Iran would not be intimidated, that a near-miss would be answered with elevation rather than retreat.

The November 29 double strike is the single clearest demonstration of the effort’s organizational depth. To hit two targets in two locations within minutes requires two complete teams, each with its own surveillance, its own approach plan, its own escape route, and its own assembled device, all operating to a synchronized clock inside a hostile capital under a security service that knew it was being hunted. This is not the work of a handful of freelancers. It is the work of an apparatus with reliable in-country infrastructure, recruited local assets, and the logistical capacity to keep weapons and surveillance running for the months of preparation each target required. The double strike was, in effect, the operation showing its hand: not a series of lucky opportunist hits, but a standing capability.

The Iranian state’s reaction to the 2010 wave shaped everything that followed. Tehran did three things at once. It blamed Israel and the United States publicly and loudly, turning each funeral into a propaganda event. It tightened the physical security of surviving scientists, adding bodyguards, armored vehicles, and varied routes. And it began, more quietly, the long internal hunt for the local networks that made the strikes possible, a counterintelligence campaign that would produce arrests, confessions broadcast on state television, and executions. Some of those confessions were almost certainly coerced, and some of the people executed may have had nothing to do with the killings. But the broader point stands. The 2010 wave forced Iran to spend enormous resources defending its scientists and purging its institutions, and that defensive cost, the time and money and attention diverted into protection, was itself one of the effort’s intended effects. A program spending its energy guarding its people is a program not spending that energy enriching uranium. This logic of imposed defensive cost is one that the Indian campaign in Pakistan would apply with equal deliberation.

Worth noting, too, is what the 2010 wave did to the relationship between Iranian scientists and their own state. A program that loses people to assassination develops an internal climate of suspicion. Every local employee becomes a potential informant in the eyes of the security service, and every security measure becomes a daily reminder to the scientists themselves that their work has made them prey. Morale is not a soft, unmeasurable thing in a technical enterprise; it determines whether senior people stay, whether junior people are willing to rise, and whether the institution can recruit at all. The 2010 wave injected a corrosive distrust into the Iranian program that no IAEA report would ever record but that anyone running the program would have felt every working day. The corrosion of trust, as much as the loss of individuals, was the wave’s lasting damage. There is a cruel efficiency to this. An adversary who kills two scientists has removed two scientists, a finite and replaceable loss. An adversary who kills two scientists in a manner that makes every survivor wonder which colleague, which guard, which contractor passed the information has damaged something far larger and far harder to repair, because trust, once broken inside an institution, is not restored by hiring. The security services that should have been protecting the program instead became, in the eyes of the scientists, a second source of menace, interrogating their families and treating their associations as suspect. A program attacked this way turns inward and spends its energy policing itself, which is exactly the outcome the attacker wants.

The 2011 and 2012 Killings

The effort did not pause to let Iran recover. In July 2011 it claimed Darioush Rezaeinejad, a young electrical engineer in his thirties, shot dead by gunmen on a motorcycle outside his home in eastern Tehran as he waited with his wife near their daughter’s kindergarten. Rezaeinejad’s death marked a small but meaningful shift in method. Where the 2010 strikes had used magnetic bombs, this was a straightforward shooting, gunmen on a motorcycle firing at close range and vanishing into traffic. The shift suggests a campaign with more than one technique in its repertoire, choosing the method that fit the specific target, the specific location, and the specific window. Rezaeinejad’s exact role has been debated. Iranian and Western accounts have linked him variously to high-voltage switches relevant to detonation systems and to other elements of the program; his family insisted he was an ordinary academic. The ambiguity, again, is part of the pattern.

Then came January 2012 and the killing described at the opening of this account. Mostafa Ahmadi-Roshan, the chemical engineer and Natanz deputy director, died when a magnetic bomb was attached to his Peugeot near Gol Nabi Street during the morning rush. His driver was also killed. By this point the operation had a rhythm that Iranian society could feel. Five connected deaths in five years, the survivor Abbasi promoted in defiance, the method now familiar enough that any scientist stuck in Tehran traffic had reason to watch his mirrors. Iran’s foreign ministry sent a formal letter to the United Nations. The Iranian parliament observed moments of silence. Funerals filled with chants. And the operation, having made its point with the Ahmadi-Roshan killing, went quiet again, as if a phase had closed.

It is worth dwelling on what that five-year sequence accomplished as a piece of psychological engineering, because the psychological effect is as important as the body count and is frequently underestimated. The effort did not need to kill every scientist in Iran. It needed only to kill enough of them, visibly enough, that every remaining scientist understood the program as a personal hazard. A metallurgist deciding whether to accept a senior post at Natanz in 2013 was not weighing salary against prestige. He was weighing salary against the morning commute, against the motorcycle in his mirror, against the kindergarten where Rezaeinejad had waited. Recruitment slows under that pressure. Senior people request transfers. Families apply quiet pressure to get out. None of this shows up in an IAEA report, but all of it degrades a program, and the degradation is cumulative. The Iranian operation understood that the target was not only the individual scientist but the willingness of the entire scientific community to keep doing the work. India’s planners absorbed the same insight, and this series has traced how the targeting of predictable daily routines in Pakistan, the prayer times and the school runs, performs the identical function: it makes an entire class of people feel hunted, so that the effort’s reach is multiplied far beyond the number it actually kills.

The Stuxnet Track: Sabotage Running in Parallel

The assassinations were never the whole story. Running alongside the killings, on a separate track, was an effort to attack the centrifuges themselves, and this second track produced what is still the most famous cyberweapon ever deployed.

Sometime in the second half of the 2000s, computer security researchers began noticing an extraordinarily sophisticated piece of malware in the wild. It became known as Stuxnet. Unlike ordinary malware, which steals data or holds files for ransom, Stuxnet was designed to do nothing on almost every machine it infected. It would spread, quietly, through Windows systems, doing no harm, until it found one very specific configuration: an industrial control system of a particular make, connected to frequency converters of a particular type, spinning equipment at the rotational speeds characteristic of uranium enrichment centrifuges. Only then did it act. It would subtly alter the speed of the centrifuges, spinning them too fast and then too slow, stressing the delicate rotors toward failure, while simultaneously feeding the plant’s operators falsified readings showing everything as normal. The centrifuges would break, and the engineers watching the monitors would see nothing wrong.

Stuxnet has never been formally claimed, but the consensus of security researchers and investigative journalists holds that it was the product of a joint American and Israeli effort, reportedly code-named Olympic Games, and that its target was the enrichment hall at Natanz. The reporting of David Sanger of The New York Times, who documented the program in detail, describes a deliberate dual-track logic, in which a cyberweapon and a human-elimination campaign were run as complementary instruments aimed at the same program. The two tracks were not redundant. They attacked different layers. The malware degraded the machines. The assassinations degraded the people who designed, ran, and repaired the machines. A program could perhaps absorb either blow alone. Absorbing both at once, while also under sanctions and inspection, was a different order of difficulty.

There is a subtler effect of Stuxnet that the dramatic headline of a digital weapon tends to obscure, and it deserves attention because it is the part most relevant to a study of covert pressure. When centrifuges began failing at Natanz, the Iranian engineers did not immediately know why. They could not be sure whether they were looking at a manufacturing defect, a metallurgical flaw in the rotors, an operator error, sabotage of the physical supply chain, or something stranger. For months, the program chased the wrong explanations, replacing hardware that was not faulty, second-guessing procedures that were sound, and losing confidence in instruments that were lying to them. Stuxnet did not only break machines. It broke the program’s trust in its own ability to know what was happening inside its own facility, and a technical enterprise that cannot trust its instruments is an enterprise that slows to a crawl. Doubt, deliberately injected, is one of the most efficient weapons a covert effort can deploy, because the adversary then spends his own resources fighting an enemy he cannot locate. The parallel to the assassination track is exact. A magnetic bomb on a Tehran street does not only kill a scientist; it makes every other scientist distrust his own commute, his own car, his own neighborhood. A worm in a control system does not only break a centrifuge; it makes every engineer distrust his own readings. In both cases the lasting damage is not the physical loss but the doubt, and doubt is cheap to create and expensive to dispel. A state that understands this will design its covert operations to maximize uncertainty in the adversary’s mind, because uncertainty multiplies the effect of every individual strike.

What makes the Stuxnet track relevant to a series about India’s shadow war is the principle it reveals, not the code. The principle is that a covert effort against an adversary’s strategic capability is most effective when it attacks multiple layers at the same time and forces the adversary to defend all of them simultaneously. Iran, after Stuxnet and the assassinations, had to harden its computer networks, air-gap its industrial controls, screen its software supply chain, protect its scientists with bodyguards and armored cars, purge its institutions of suspected local collaborators, and do all of it under the constant awareness that it did not know where the next blow would land or in what form. The defensive burden was total. Every layer of the program became a layer that had to be guarded, and guarding is expensive, slow, and demoralizing. India’s shadow war in Pakistan has shown a similar instinct for multi-layered pressure, combining the elimination of operatives with financial disruption, diplomatic exposure, and the conventional military option, and the modus operandi of covert eliminations examined elsewhere in this series reads, in its underlying logic, like a descendant of the Iranian dual-track model.

Stuxnet also carried a warning that India’s planners would have been wise to absorb. The weapon eventually escaped. It spread beyond its intended target, infected machines around the world, and was discovered, dissected, and published precisely because it had gone where it was not meant to go. A covert instrument, once loose, can become public, and a public covert instrument is no longer covert. The lesson is that even the most disciplined campaign carries a risk of exposure that grows with every operation, and that exposure can convert a deniable success into an undeniable diplomatic problem. The Iranian campaign as a whole would eventually run into exactly that wall.

The Fakhrizadeh Operation: A Robot on the Absard Road

For most of the 2010s the assassination track was quiet. The 2015 nuclear agreement between Iran and six world powers froze much of the enrichment program in exchange for sanctions relief, and during the life of that agreement the operation appeared dormant. Then the United States withdrew from the agreement in 2018, Iran began rebuilding its enrichment capacity, and the operation returned for its most ambitious operation yet.

On November 27, 2020, Mohsen Fakhrizadeh was traveling by car on a rural road near the town of Absard, in the hills east of Tehran, when he was killed in an ambush. Fakhrizadeh was not an ordinary scientist. He was widely described as the central figure of Iran’s nuclear weapons research, the long-time head of an effort sometimes called the Amad Plan, and the man Israeli officials had singled out by name in public presentations as the person to watch. If the operation had a single highest-value target, it was Fakhrizadeh, and his protection reflected that. He traveled with a security detail. His movements were guarded. He was, by every measure, the hardest target the effort had ever attempted.

The method used to kill him was, according to detailed reporting based on Israeli sources, a remote-controlled weapon. Rather than placing a team of gunmen physically on the road, where they could be caught or killed in the firefight with Fakhrizadeh’s bodyguards, the operation reportedly used a machine gun mounted on a parked vehicle, aimed and fired remotely, possibly with software assistance to compensate for the delay in the satellite link and for the movement of the target’s car. The weapon fired, killed Fakhrizadeh, reportedly left his wife beside him unharmed by the precision of the targeting, and the vehicle that carried it was then destroyed. No operative had to be physically present at the moment of the kill.

The Fakhrizadeh operation is the effort’s clearest glimpse of its own future, and the implication runs directly into the present concerns of this series. The whole logic of the motorcycle assassination, in Tehran and later in Pakistan, is the management of risk to the operatives: get close, strike, and vanish into traffic before anyone can react. A remote weapon removes the operative from the equation entirely. There is no rider to catch, no shooter to interrogate, no human being whose capture could unravel the network or produce a confession. The trade-off is that a remote weapon requires far more advance preparation, sophisticated engineering, and the ability to emplace and conceal a complex device in advance. The Fakhrizadeh killing suggested that the most advanced practitioners of targeted killing were moving toward a model in which the assassin is a machine and the human role recedes to planning and command. Any state running a campaign of targeted eliminations, including India, is now operating in a world where that option exists and has been proven. The motorcycle is not obsolete, but it is no longer the only tool, and a campaign that fails to think about the remote-weapon model is a campaign reasoning with yesterday’s catalog.

Any state running a program of targeted eliminations, India included, now operates in a world where the remote-weapon option exists and has been proven on a hard target. The motorcycle has not become obsolete, but it is no longer the only available tool. Equally, the remote weapon raises the entry cost steeply. It demands engineering depth, smuggling capacity, and the ability to position and conceal a heavy, complex device for an extended period in hostile territory. A state without those resources will keep relying on the motorcycle and the handgun, which remain effective precisely because they are cheap, simple, and built from components available anywhere. The Fakhrizadeh case therefore did not make older methods irrelevant. It widened the menu, and it signaled that the most capable actors will increasingly choose the option that exposes no human operative at all. For India, this matters in a specific way. The Pakistani killings have so far relied overwhelmingly on the human-operative model, the two riders and the handgun, and that model carries an irreducible risk: an operative can be caught, and a caught operative can be interrogated, paraded, and used to convert a deniable program into an attributed one. The remote weapon removes that risk entirely. If India’s planners are studying the Iranian precedent as closely as the methodological resemblance suggests, the Fakhrizadeh operation is the part of the record they will have examined hardest, because it points toward a future in which the single greatest vulnerability of a covert killing program, the human being who pulls the trigger, is engineered out of the equation.

Iran’s response to the Fakhrizadeh killing followed the now-familiar script, but with sharper anguish, because this target was irreplaceable in a way the others had not been. Tehran promised revenge, blamed Israel directly, and used the killing to argue at home that negotiation with the West was futile because the West, or its regional partner, would always reach for the gun. The killing also hardened a faction inside Iran that had always argued the program should be pushed faster and buried deeper, the precise opposite of what the effort’s planners presumably intended, and that paradox sits at the center of the effectiveness debate this account turns to shortly.

From Covert to Overt: The 2025 Escalation

For two decades the effort against Iran’s scientists obeyed one iron rule: deniability. The deaths were attributed but not claimed, the methods were chosen for their ambiguity, and the official Israeli posture was silence. In June 2025 that rule was discarded in the most dramatic way imaginable.

On June 13, 2025, Israel launched a large-scale military operation against Iran, named Operation Rising Lion, opening what became known as the Twelve-Day War. The Israeli Air Force struck nuclear facilities, military bases, and air defenses across the country, and in the opening hours the strikes deliberately targeted the homes of nuclear scientists. By the Israeli military’s own account, the operation killed nine senior scientists and specialists associated with the nuclear weapons effort in its first wave. The precision of the strikes on private residences, in the middle of the night, made clear that Israeli intelligence had maintained a detailed and current map of where each scientist lived. The effort that had once placed a bomb on a motorcycle now placed a bomb from an aircraft, and it did so under the national flag, announced by the prime minister, framed not as a deniable accident but as a declared act of war.

The shift from covert to overt is the single most analytically important development in the entire history of the effort, and it deserves to be understood clearly rather than treated as a mere escalation. For twenty years the assassination of Iranian scientists had been a deniable instrument precisely because deniability kept the operation below the threshold of war. A magnetic bomb on a Tehran street could be condemned, investigated, and mourned, but it did not, by itself, oblige Iran to mount a full military response, because Iran could not formally prove who had done it and a formal proof is what a declaration of war conventionally requires. The covert method was, in this sense, a method of escalation control. It allowed Israel to degrade the program while keeping the conflict contained.

The 2025 airstrikes abandoned that control deliberately. By killing scientists openly, from the air, under the flag, Israel converted the effort from a deniable nuisance into a casus belli, and Iran responded with hundreds of ballistic missiles and a wider war that drew in the United States. Whether that trade was worth it is a judgment that depends on what Israel believed the alternative to be, and the most coherent reading is that Israeli planners had concluded the covert track had reached the limit of what it could achieve. Killing one scientist every couple of years could harass a program. It could not stop one that was, by mid-2025, assessed to be closer than ever to a weapons capability. When a covert effort’s pace can no longer match the speed of the threat, the state running it faces a hard choice: accept the threat or escalate openly. Israel chose to escalate.

Pause on the geometry of that choice, because it is the geometry every state running a covert program eventually confronts. A deniable operation is cheap in the currency that matters most, which is the risk of full-scale war. It is also slow, because deniability constrains tempo; a state cannot kill scientists every week without the pattern becoming so blatant that deniability collapses on its own. Slow and safe is an acceptable trade only as long as the adversary’s progress is also slow. The moment the adversary accelerates, the covert program’s central limitation, its low tempo, becomes a fatal flaw, because a tool that removes one specialist a year cannot keep pace with a program graduating specialists by the dozen. At that point the state must either accept the adversary’s progress or trade away the safety of deniability for the speed of open force. Israel reached that fork in 2025 and chose speed. The decision was not reckless; it was the logical terminus of a covert program that had been outrun by its target. But the terminus is the point, and it is the warning. A covert effort does not run forever. It runs until the threat outpaces it, and then its operator faces a choice that is genuinely worse than the choices available at the start, because by then the program has already spent years of deniability, hardened the adversary’s resolve, and narrowed the diplomatic options. The covert phase, in other words, does not only buy time. It also shapes, and often worsens, the situation the state will face when the covert phase ends. That is the hidden cost no ledger of successful strikes will show.

This is the precise dilemma that this series has examined in the context of India and Pakistan, and the parallel is exact enough to be uncomfortable. India’s shadow war is, like the covert phase of the Iranian effort, an instrument of escalation control. It allows New Delhi to impose costs on Pakistan-based networks without crossing the threshold that would oblige a conventional military response. But the Iranian case demonstrates that the covert instrument has a ceiling. If the threat outpaces what deniable operations can suppress, the temptation to escalate openly becomes very strong, and the comparison to the open military dimension of the India-Pakistan confrontation is one this series treats as central. The covert and the overt are not separate doctrines. They are points on a single ladder, and the Iranian effort is the clearest case study the world has of a state climbing from one rung to the next.

Key Figures

The story is best understood through the individuals it touched, and not all of them were scientists. A complete picture requires looking at the men who died, the man who survived, and the institutional figures who shaped the effort on both sides.

Mohsen Fakhrizadeh

Fakhrizadeh was the effort’s defining target and, in death, its most consequential. A physicist by training and a long-serving officer in Iran’s Revolutionary Guard structure, he was widely identified as the organizing intelligence behind Iran’s weapons-relevant research, the figure who held the dispersed pieces of the effort together. Israeli officials had named him publicly years before his death, an unusual move that effectively marked him. His killing in November 2020 removed a manager whose value lay not in any single technical skill but in his irreplaceable institutional knowledge of how the entire enterprise fit together. Iran treated his death as a national wound and, revealingly, responded by passing legislation to expand enrichment, the opposite of capitulation. Fakhrizadeh’s case is the strongest evidence for both sides of the effectiveness debate: he was important enough that his loss genuinely hurt the program, and his killing genuinely hardened Iran’s resolve. The naming of him in public by Israeli officials before his death is itself worth study, because it inverted the usual logic of deniability. To name a target in advance is to announce intent, and announcing intent is normally something a covert program avoids. The decision suggests that Israeli planners had concluded Fakhrizadeh was important enough that the deterrent value of marking him publicly outweighed the operational cost of warning him.

Fereydoon Abbasi-Davani

Abbasi is the effort’s most instructive survivor. A physicist deeply embedded in the program, he lived through the November 2010 attack because he sensed the danger and reacted in the seconds that mattered. His subsequent appointment to lead Iran’s atomic energy organization turned a failed assassination into a propaganda asset for Tehran, a living demonstration that the campaign could be defied. Abbasi’s survival also carries an operational lesson that this series has noted in the Pakistani context: a target who knows he is being hunted, and who behaves accordingly, is dramatically harder to kill than one who does not. The campaign’s later move toward remote weapons can be read partly as a response to the Abbasi problem, an attempt to remove the margin of reaction time that an alert target can exploit.

Majid Shahriari and Mostafa Ahmadi-Roshan

Shahriari, killed in the November 2010 double strike, and Ahmadi-Roshan, killed in January 2012, represent the campaign’s core category of target: mid-to-senior technical specialists with direct, demonstrable connections to enrichment work. Neither was as singular as Fakhrizadeh, but both were genuine nodes of expertise, and their killings degraded real capability rather than merely sending a message. Ahmadi-Roshan in particular, as a deputy director at Natanz, sat close enough to the enrichment effort that his loss touched the program’s daily operation. Together they define what the operation was, at its center, actually doing: not theater, but the deliberate subtraction of trained people from a hard problem. Neither man’s name is widely known outside specialist circles, and that obscurity is part of the point. An effort designed to degrade capability rather than to send a message will concentrate on exactly these mid-tier specialists, the people whose loss is felt inside the laboratory long before it is felt in the headlines.

Masoud Alimohammadi and Darioush Rezaeinejad

Alimohammadi, the particle physicist killed in January 2010, and Rezaeinejad, the young electrical engineer killed in July 2011, represent the campaign’s contested margin. Neither had a publicly established role in enrichment, and in both cases the families and some independent analysts argued the men were academics rather than weaponeers. Their deaths raise the campaign’s hardest question: how confident was the targeting, and how many of the dead were chosen on intelligence that was strong rather than merely suggestive. An honest account cannot resolve this, but it must name it. An effort that kills on imperfect intelligence will, over a long enough run, kill some people who did not belong on the list, and the Iranian operation almost certainly did. The contested cases also complicate the moral defense of the whole enterprise, because a defense that rests on the certain guilt of the target is only as strong as the intelligence behind the targeting, and intelligence is never certain. This is a vulnerability India’s shadow war shares in full measure.

David Sanger and Ronen Bergman

No account of the campaign would exist without the journalists and authors who reconstructed it from sources that the states involved would never confirm. David Sanger’s reporting documented the dual-track logic that linked Stuxnet to the assassinations and established the existence of the cyber program in public. Aaron Klein and other writers on Israeli intelligence operations reconstructed the operational methodology. Their work matters here for a methodological reason. A covert effort, by design, leaves no official record, and everything the public knows about both the Iranian and the Pakistani campaigns is the product of investigative reconstruction rather than government disclosure. That should make every confident claim, including the claims in this account, appropriately provisional.

Consequences and Impact: Did the Campaign Work?

The central question, stripped of drama, is simple. After fifteen years of assassinations and a decade and a half of sabotage, was Iran’s nuclear program meaningfully delayed? And the honest answer is that the evidence cuts both ways, which is exactly why the question is worth taking seriously rather than answering with a slogan.

Arguments that the effort worked rest on the concept of cumulative friction. Each killing removed a trained specialist whose replacement took years. Each killing forced the program to divert money and attention into protecting its remaining people. Stuxnet physically destroyed centrifuges and, just as importantly, injected a paralyzing distrust into the program’s relationship with its own instruments, because for months Iranian engineers could not be sure whether a malfunctioning cascade was a hardware fault or a hidden attack. The recruitment of new scientists slowed under the visible hazard of the work. Taken together, the argument runs, these effects bought years. Iran in the 2010s advanced more slowly than an unmolested program would have, and in a strategic competition where the goal is to delay an adversary until politics or sanctions or regime change can intervene, buying years is a real achievement. By this measure the campaign succeeded, not by stopping the program but by slowing it.

The case that the campaign failed rests on a single stubborn fact: Iran’s program did not stop, and by mid-2025 it was assessed to be closer to a weapons capability than at any point in its history. The IAEA reported that summer that Iran had accumulated enough enriched uranium to produce material for multiple weapons. If the purpose of fifteen years of killing was to prevent Iran from reaching the nuclear threshold, then fifteen years of killing did not achieve the purpose, because Israel ultimately judged that only open war could address the threat the covert effort had failed to eliminate. Worse, the critics argue, the campaign may have been counterproductive in a way the friction argument ignores. Every assassination strengthened the Iranian faction that argued the program must be accelerated, hardened, and buried deeper. Every killing converted the program from a bureaucratic enterprise into a national cause defended by martyrs. A program attacked this way does not necessarily slow down. It can speed up, dig in, and disperse, becoming harder to monitor and harder to stop than it would have been if left alone. The replacement of killed scientists by new recruits, radicalized by the deaths of their predecessors, is a dynamic the friction model does not capture.

Consider the arithmetic of replacement, because it is the hidden engine of the entire effort. When a senior centrifuge engineer is killed, his employer does not simply hire another senior centrifuge engineer, because senior centrifuge engineers are not available on a labor market. The institution must instead promote someone less experienced, who will make the mistakes that experience would have prevented, and then train a replacement from a lower tier, a process that takes years and during which the program runs on thinner expertise. Each killing therefore imposes a cost that compounds over time and across the institution, not a single subtraction but a cascade of demotions in capability. A program that loses five senior people over five years does not lose five units of capability. It loses the five, plus the years of reduced output while replacements climb the curve, plus the judgment those five would have applied to problems their juniors will now mishandle, plus the recruits who never joined because the work became visibly lethal. The true cost is a multiple of the body count, and it is invisible to anyone counting only funerals.

Pakistani militants and Iranian scientists are, on the surface, entirely different categories of person, and a reader might reasonably object that the comparison strains. The objection has force, and this account returns to it at length in the analytical section. On the specific question of psychological reach, however, the two categories behave identically. A militant commander who learns that three of his peers have been shot at close range while leaving a mosque changes his behavior in exactly the way a physicist who learns that three of his peers have been bombed in traffic changes his. He varies his routine, he trusts fewer people, he moves his family, he wonders which of his associates has been turned. The terror of being hunted is not specific to the kind of work the hunted person does. It is specific to the certainty that a competent, patient, well-resourced organization has placed your name on a list, and that certainty does the same corrosive work in Lahore that it does in Tehran. This is why the two efforts, despite targeting such different people, produce such similar second-order effects. Both shrink the pool of people willing to do the dangerous work, both force the surviving members into defensive routines that reduce their productivity, and both convert an organization’s internal trust into an internal liability. The method is not merely transferable between the two theaters; the human response to the method is essentially constant.

The most defensible assessment is that both things are true at once, and that the discomfort of holding them together is the actual lesson. The campaign almost certainly slowed the program’s pace in the short and medium term. It also almost certainly hardened the program’s political foundation and pushed it toward deeper concealment in the long term. It bought time and it raised the eventual cost of using that time. Whether the trade was worth it depends entirely on what Israel did with the years it purchased, and the answer to that turned out to be: escalate to open war in 2025. A covert effort that succeeds in buying a decade, only for the decade to end in the very war the campaign was meant to avoid, occupies an ambiguous place in any honest ledger.

There is one more consequence that belongs in the ledger, and it concerns the price paid by people who were not the target. The 2010 and 2012 magnetic bombs killed not only the scientists but, in at least one case, a driver, and they were detonated in dense morning traffic where any bystander could have been caught by the blast or by flying debris. The remote weapon used against Fakhrizadeh was reportedly precise enough to spare his wife in the same car, which is genuinely a mark of operational care, but precision of that order was not available in the earlier years and was not the norm. An operation conducted on crowded streets accepts, as a structural matter, a risk to uninvolved people, and that risk is part of the true cost even when it does not materialize. Any honest comparison with India’s operations in Pakistan, which also unfold on crowded streets and outside mosques, must carry the same line item. This is not an argument that either effort is uniquely careless; both appear, on the available evidence, to have tried to limit harm to bystanders, and the trend toward remote and precision methods reduces the danger further. But an operation that places explosives on cars in rush-hour traffic, or that fires weapons in public streets, cannot honestly claim that the risk to the uninvolved is zero. The risk is real, it is accepted deliberately as the price of operating in the adversary’s cities, and a sober accounting names it rather than burying it beneath the cleaner language of precision and proportionality.

This ambiguity is the single most important thing the Iranian case offers India, and it is worth stating without hedging. A campaign of targeted killing can degrade an adversary’s capability. It is far less clear that it can eliminate that capability, and there is real evidence that it can entrench the adversary’s resolve. India’s shadow war against Pakistan-based networks is frequently assessed in this series and elsewhere on its tactical success, the number of operatives removed, the disruption imposed. The Iranian precedent is a warning that tactical success and strategic success are different things, and that a campaign can run for fifteen years, kill its targets with professional reliability, and still arrive at a strategic situation no better, and possibly worse, than where it began.

Analytical Debate: The Iran Campaign Measured Against India’s Shadow War

The reason this account belongs on a site devoted to India’s shadow war is that the Iranian effort is its closest functioning analogue, and the comparison is most useful when it is made systematically rather than impressionistically. What follows places the two campaigns side by side across five dimensions: the type of target, the method of killing, the level of attribution, the management of consequences, and the strategic effect. Each dimension reveals where the campaigns converge and, just as importantly, where they diverge.

Target type is the first dimension, and here the campaigns diverge sharply, in a way that matters morally and legally. India’s shadow war in Pakistan targets men with documented histories of organizing or facilitating terrorist violence: operatives of groups responsible for mass-casualty attacks, men named in charge sheets and on international sanctions lists. Whatever one thinks of extrajudicial killing, these are individuals against whom a state can articulate a concrete grievance grounded in past acts of violence. The Iranian campaign targeted scientists and engineers, men whose offense was technical competence in service of a program their own government insisted was lawful and civilian. The Iranian target had, in the conventional sense, no blood on his hands. This is the campaigns’ deepest moral asymmetry, and it cuts in India’s favor: it is easier to defend the killing of a man who organized a massacre than the killing of a man who understands centrifuge metallurgy. The Iranian campaign’s defenders must reach for a preemptive logic, the future weapon, that India’s shadow war does not strictly require.

Method is the second dimension, and here the campaigns converge almost completely, which is the strongest single piece of evidence for doctrinal transfer. Both campaigns built their signature around the motorcycle: the two-rider configuration, the magnetic charge or the close-range shooting, the escape into the traffic of a dense city. Both selected the predictable commute as the killing window. Both, in their most recent phases, showed movement toward more advanced methods, the Iranian effort in the remote weapon used against Fakhrizadeh and the open airstrikes of 2025. The methodological fingerprint is so similar that the Pakistani motorcycle assassinations could be described, without much exaggeration, as the Tehran method relocated to Karachi. Either India studied the Iranian template directly, or both campaigns, solving the same tactical problem, arrived at the same solution. The practical consequence is identical regardless of which explanation is true.

Attribution is the third dimension, and here the campaigns again converge, but with an instructive difference of degree. Both operate in the gap between attribution and acknowledgment. Everyone knows who is responsible; no one formally admits it. The Iranian campaign, however, eventually crossed out of that gap in 2025, when Israel claimed its airstrikes openly. India’s shadow war, as of the period this series covers, has not crossed that line. New Delhi continues to deny involvement in the Pakistani killings, characterizing the allegations as malicious propaganda. The Iranian case therefore shows India both the value of the deniability India still maintains and the conditions under which that deniability might eventually be discarded. As long as the covert instrument can keep pace with the threat, deniability is worth preserving. When it cannot, the Iranian precedent suggests, the pressure to acknowledge and escalate becomes hard to resist.

Consequence management is the fourth dimension, and here the campaigns diverge in a way that should worry Indian planners. Israel, in running the Iranian effort, faced a target state that could retaliate substantially. Iran answered the 2025 escalation with hundreds of ballistic missiles and a regional war. The covert phase had been calibrated precisely to avoid triggering that response, and the moment the campaign went overt, the response arrived. India’s operations in Pakistan faces a target state that is also nuclear-armed and also capable of retaliation, conventional and unconventional. The Iranian case is a demonstration of what happens when the careful management of consequences fails: the conflict that the covert method was designed to prevent arrives anyway, and at full scale. Any assessment of India’s shadow war that treats Pakistani retaliation as a remote possibility should sit with the Iranian example, in which the retaliation was not remote at all once the threshold was crossed.

Strategic effect is the fifth dimension, and here the campaigns share the same fundamental uncertainty, the uncertainty this account has already named. Neither campaign has clearly achieved its ultimate strategic objective. The Iranian campaign did not stop Iran’s program; it slowed it and then handed off to open war. India’s operations has not ended the threat from Pakistan-based militancy; it has imposed costs and disrupted networks without eliminating the underlying sanctuary. Both campaigns demonstrate that targeted killing is a real instrument with real tactical effects, and both demonstrate that the instrument has a strategic ceiling. The honest verdict on the comparison is that India’s operations is morally easier to defend, methodologically nearly identical, and strategically subject to the same hard limit: it can degrade, harass, and delay, but the evidence that it can deliver a decisive, lasting result is, in both cases, thin.

A sixth comparison sits outside the formal five dimensions but deserves a place in any honest analysis, because it concerns the question of which model has handled exposure better. The Iranian effort was, for most of its life, remarkably well concealed in the sense that no operative was paraded on Iranian state television as an Israeli agent in the way the planners would have feared. Yet the cyber track exposed itself when Stuxnet escaped, and the assassination track was so consistent in method that deniability eroded through sheer repetition long before 2025. India’s operations have so far avoided a Stuxnet-style self-exposure, but they too have accumulated a pattern visible enough that a major Western newspaper was able to publish an investigation attributing the Pakistani killings to Indian intelligence. The lesson common to both is that deniability is not a permanent property of a covert program. It is a depleting resource. Every operation spends a little of it, and a program that runs long enough will eventually run the account down to zero, at which point the state must decide whether to stop or to escalate. This reframes how a covert program should be judged. A single operation can be assessed on whether it succeeded and whether it stayed deniable. A long program cannot, because the very repetition that makes it a program also guarantees the erosion of the deniability that made it attractive. The honest question for India is therefore not whether any individual operation in Pakistan can be kept deniable, which it probably can, but how much deniability the cumulative program has already spent, and how much remains before New Delhi faces the same fork Israel reached in 2025. The Iranian experience suggests that fork arrives sooner than the program’s operators expect, and that they are rarely ready for it when it does.

This five-dimension comparison is the analytical core of why the Iranian effort deserves study by anyone trying to understand India’s shadow war. It is not a perfect mirror. The target type differs, and that difference matters. But on method, on the attribution gap, on the escalation ladder, and on the strategic ceiling, the two campaigns rhyme so closely that the Iranian experience functions as something close to a forecast. India is, in important respects, running the Tehran campaign a decade later, against a different adversary, and the Iranian outcome, fifteen years of professional killing followed by the open war the killing was meant to prevent, is a possibility that Indian strategy cannot responsibly ignore.

Why It Still Matters

The campaign against Iran’s nuclear scientists is, on its surface, a story about one specific rivalry in one specific region. Read more carefully, it is a story about a method that has escaped its origin and become a general tool of statecraft, and that is why it still matters far beyond the Israel-Iran file.

Normalization is the first reason the story still matters. The sustained operation treated the killing of scientists as a usable instrument of policy. Before the Iranian campaign, the targeted assassination of an adversary’s weapons researchers, men in no uniform and accused of no violence, was close to taboo. The Iranian campaign, conducted over fifteen years with no decisive international consequence, eroded that taboo. A state contemplating a similar campaign today operates in a normative environment that the Tehran killings helped create, and that environment is more permissive than the one that existed in 2007. Norms erode through precedent, and the Iranian campaign is now a precedent that other states can and do cite, explicitly or implicitly, when they reach for the same tool.

The escalation ladder is the second reason it matters, and it is the reason most relevant to this series. The Iranian story is the most complete available case study of the covert-to-overt progression. Most covert efforts are studied only in their covert phase, because most do not visibly escalate. The Iranian campaign did. It ran for two decades as a deniable operation and then, in 2025, climbed openly into war. That makes it uniquely valuable for understanding what covert efforts are actually for, and what their limits are. The lesson is that a covert campaign is rarely an end state. It is a phase, an instrument for buying time and managing escalation, and the time it buys eventually runs out. Any state running such a campaign, including India, is somewhere on that ladder, and the Iranian case is the clearest map of where the higher rungs lead.

Effectiveness, the unresolved question, is the third reason, and the Iranian experience poses it more sharply than any other case. The campaign was, tactically, a sustained professional success. Its targets died on schedule, its methods worked, its operatives were rarely caught. And it still did not achieve its strategic purpose. That gap, between flawless tactics and an ambiguous strategic result, is the most important thing the campaign teaches. It is a warning against the seductive assumption that a campaign which is killing its targets must therefore be winning. The legal and ethical debate around targeted killing and the comparison of covert killing with other counter-capability methods such as drone warfare both ultimately circle the same question that the Iranian campaign forces into the open: not whether targeted killing can be done, which is settled, but whether doing it accomplishes what its planners believe it will.

Beyond the two countries directly involved, the Iranian story matters because it has quietly rewritten one of the rules of the international system. For most of the postwar era, the assassination of an adversary’s officials and specialists in peacetime was treated as conduct that even hostile states mostly avoided, not out of sentiment but because the precedent was understood to be dangerous to everyone. A world in which states routinely kill each other’s scientists is a world in which every state’s scientists are at risk, and that mutual vulnerability sustained a rough restraint. The sustained operation against Iran, conducted for fifteen years with no decisive penalty, tested that restraint and found it weaker than assumed. Each unanswered killing made the next one a little more thinkable, not only for Israel but for every government watching. The erosion is incremental and easy to miss, but it is real, and it is the kind of change that becomes visible only in retrospect, when a practice that was once exceptional has quietly become ordinary. India’s shadow war is one data point in that larger drift, and the drift is the genuine long-term consequence. The individual scientists are dead, the individual militants are dead, and those are finite facts. The shift in what states feel entitled to do to one another is not finite. It compounds, and the Iranian operation, more than any other single case, is the one that moved it furthest.

For India, the Iranian campaign is the mirror held up to its own shadow war, and the reflection is sobering rather than flattering. It shows a campaign that did everything India’s campaign is doing, did it with skill, did it for longer, and arrived at an outcome no one in Tel Aviv in 2007 would have chosen: the very war the killing was meant to prevent. India’s planners have built an impressive covert instrument, and this series has documented its operations in detail. The Iranian precedent does not say that instrument is worthless. It says the instrument has a ceiling, that the ceiling is reached sooner than its users expect, and that the most important strategic decisions are not about how to run the covert campaign but about what to do when the covert campaign reaches its limit. That is the question the Tehran killings ultimately pose, and it is a question India has not yet had to answer, but on the evidence of Iran, eventually will.

Frequently Asked Questions

Q: How many Iranian nuclear scientists were assassinated?

The most commonly cited covert phase of the campaign, running from roughly 2007 to 2020, is associated with the deaths of five to six men connected to the nuclear program: Ardeshir Hosseinpour in 2007, whose killing is disputed, Masoud Alimohammadi in January 2010, Majid Shahriari in November 2010, Darioush Rezaeinejad in July 2011, Mostafa Ahmadi-Roshan in January 2012, and Mohsen Fakhrizadeh in November 2020. A further attack in November 2010 targeted Fereydoon Abbasi-Davani, who survived. The 2025 airstrikes of Operation Rising Lion then killed, by Israel’s own account, nine more senior scientists and specialists in their opening wave, which means the total number of program-linked specialists killed across the covert and overt phases is well into double figures. Precise counting is complicated by the disputed status of some deaths and by the difficulty of confirming which of the 2025 airstrike victims held genuinely senior roles.

Q: What methods did Israel use against Iranian scientists?

The covert phase relied on a small set of techniques. The most common was the magnetic bomb, a shaped charge attached to a moving car by a passing motorcyclist and detonated by remote control or magnetic adhesion, which was the method used against Shahriari, Abbasi, and Ahmadi-Roshan. A second technique was the close-range shooting from a motorcycle, used against Rezaeinejad. A third was the remote-controlled weapon, reportedly a machine gun mounted on a parked vehicle and fired remotely, used against Fakhrizadeh in 2020. The 2025 phase abandoned covert methods entirely in favor of airstrikes on scientists’ homes. Running alongside the killings throughout was the Stuxnet cyberweapon, a separate sabotage track aimed at the centrifuges themselves rather than at people.

Q: Did the Iran operations use motorcycle assassins the way India does in Pakistan?

Yes, and this is the single closest point of resemblance between the two campaigns. The Iranian killings of 2010 to 2012 established the motorcycle as the signature platform: two riders, a magnetic charge or a handgun, and an escape into dense urban traffic that no pursuing vehicle could follow. The motorcycle-borne assassinations later attributed to India inside Pakistani cities use a method that is, in its essentials, indistinguishable. Whether India studied the Iranian template directly or arrived independently at the same solution to the same tactical problem cannot be proven from open sources, but the result is the same: the Tehran method and the Karachi method are functionally one method, separated only by geography and a decade.

Q: Did killing scientists actually delay Iran’s nuclear program?

This is genuinely contested. The case that it worked rests on cumulative friction: each killing removed a hard-to-replace specialist, slowed recruitment, and forced the program to spend heavily on protecting its remaining people, which together likely slowed the program’s pace through the 2010s. The case that it failed rests on the fact that Iran’s program never stopped and by 2025 was assessed to be closer to a weapons capability than ever, and that the killings may have hardened Iranian resolve and pushed the program into deeper concealment. The most defensible view is that the campaign slowed the program in the medium term while entrenching it in the long term. It bought time without delivering a decisive result, which is why Israel ultimately escalated to open war in 2025.

Q: What was the connection between Stuxnet and the assassinations?

Stuxnet and the assassinations were two tracks of a single strategy, reportedly run in coordination. Stuxnet, a cyberweapon widely attributed to a joint American and Israeli effort, attacked the centrifuges directly by subtly altering their rotation speeds to cause physical failure while hiding the damage from operators. The assassinations attacked the human layer, the scientists and engineers who designed, operated, and repaired those centrifuges. The logic of running both at once was to attack different layers of the program simultaneously and force Iran to defend all of them at the same time, multiplying the defensive burden. Investigative reporting, particularly by David Sanger, documented this dual-track design.

Q: Who was Mohsen Fakhrizadeh and why was his killing significant?

Fakhrizadeh was widely identified as the central organizing figure of Iran’s nuclear weapons research, the long-time head of the effort sometimes called the Amad Plan, and a senior officer in Iran’s Revolutionary Guard structure. Israeli officials had named him publicly years before his death. His killing in November 2020, by a reportedly remote-controlled weapon near the town of Absard, was significant for two reasons. First, his institutional knowledge of how the entire dispersed program fit together made him genuinely hard to replace. Second, the method, a remote weapon requiring no operative physically present at the kill, signaled a technological evolution in how targeted killings could be conducted, with implications for every state running such a campaign.

Q: Did Iran retaliate for the scientist assassinations?

During the covert phase, Iran’s retaliation was constrained. It blamed Israel and the United States loudly, expelled and arrested suspected local collaborators, executed some of them, tightened the security of surviving scientists, and pursued its own covert operations against Israeli and Jewish targets abroad, with mixed and often unsuccessful results. What it did not do during the covert phase was mount a full conventional military response, precisely because the deniable nature of the killings did not formally oblige one. That changed in 2025. When Israel escalated to open airstrikes, Iran responded with hundreds of ballistic missiles and a wider regional war, demonstrating exactly the retaliation that the covert method had been designed to avoid.

Q: Why did Israel target scientists rather than nuclear facilities?

Iran’s facilities were dispersed across the country and heavily hardened, much of the program buried under mountains, which made a decisive airstrike of the kind Israel had used against Iraq in 1981 unrealistic. Israeli planners reportedly concluded that if the buildings could not be destroyed, the program could still be attacked through its other vulnerabilities, and that the human layer, the irreplaceable scientists and engineers, was the softest of those vulnerabilities. Training a senior centrifuge specialist takes years, specialists are not interchangeable, and a program that loses several of them suffers in knowledge, judgment, and morale. Targeting people was, in this reasoning, a way to attack a program whose physical infrastructure was beyond conventional reach.

Q: How did Israeli intelligence identify which scientists to target?

The open-source record does not contain the targeting methodology in detail, and any answer is necessarily inferred. Israeli intelligence was reported to maintain extensive human and technical penetration of Iran, and the precision of the 2025 airstrikes on individual residences confirmed that it held a detailed and current map of where scientists lived. The selection of targets across the campaign suggests a focus on specialists with demonstrable connections to enrichment and weapons-relevant research, although the contested cases of Alimohammadi and Rezaeinejad indicate that the targeting may sometimes have rested on intelligence that was suggestive rather than conclusive. The honest position is that the campaign’s targeting was sophisticated but not infallible.

Q: Were the assassinations legal under international law?

There is no settled answer, and the legality is genuinely disputed. Defenders of the campaign argue it was an act of preemptive self-defense against an existential threat, pointing to Iran’s repeated calls for Israel’s destruction and to the weapons-relevant nature of the targeted research. Critics argue that the killings were extrajudicial assassinations on the sovereign territory of a state with which Israel was not formally at war, that the targets were not combatants in any conventional sense, and that the preemptive-self-defense logic, accepted broadly, would license the assassination of weapons researchers worldwide. The 2025 airstrikes, conducted openly, were condemned by the United Nations and many governments as a violation of international law, while others defended them as justified nonproliferation. The legal question remains unresolved.

Q: How does the Iran campaign compare to India’s shadow war in Pakistan?

The two campaigns are close analogues that converge on method and diverge on target type. They converge almost completely on technique: the motorcycle platform, the magnetic charge or close-range shooting, the predictable commute as the killing window, and movement toward more advanced methods over time. They also share the attribution gap, the position of being widely attributed but not formally claimed. They diverge on the nature of the target: India’s campaign targets men with documented histories of organizing terrorist violence, while the Iranian campaign targeted scientists accused of no violence, which makes India’s campaign easier to defend morally. Both campaigns appear to face the same strategic ceiling, capable of degrading and delaying an adversary’s capability without clearly eliminating it.

Q: Did the United States help carry out the assassinations?

The United States has consistently distanced itself from the assassination track. American officials, when the killings occurred, generally denied involvement and sometimes expressed disapproval of the method. The cyber track is different. Stuxnet is widely attributed by investigative reporting to a joint American and Israeli effort, reportedly code-named Olympic Games, so the consensus understanding is that Washington participated directly in the sabotage of the centrifuges while keeping clear of the killing of scientists. In the 2025 overt phase, the United States did not conduct the initial strikes on scientists’ homes but later joined the wider war by bombing Iranian nuclear sites, a more direct involvement than at any point in the covert phase.

Q: Why didn’t Iran’s scientists have better protection?

They did receive substantial protection after the campaign became unmistakable in 2010. Surviving scientists were given bodyguards, armored vehicles, and varied routes, and Iran mounted a major counterintelligence effort to find the local networks enabling the strikes. The difficulty is that protection has limits against a determined and well-resourced campaign. A scientist must still commute, still live somewhere, and still follow some routine, and each routine is a window. The Fakhrizadeh case showed that even a guarded senior figure with a security detail could be reached, in his case by a remote weapon that removed the bodyguards from the equation. Protection raises the cost and difficulty of an assassination; it does not make a target permanently unreachable.

Q: What was Operation Rising Lion?

Operation Rising Lion was the name Israel gave to the large-scale military operation it launched against Iran beginning on June 13, 2025, opening what became known as the Twelve-Day War. The operation struck nuclear facilities, military bases, and air defenses across Iran, and its opening wave deliberately targeted the homes of nuclear scientists, killing nine senior specialists by Israel’s own account. It marked the decisive shift in the campaign from covert, deniable assassination to open, claimed military action. Iran retaliated with hundreds of ballistic missiles, the United States later joined by bombing Iranian nuclear sites, and a ceasefire was reached after twelve days, though tensions and further strikes followed.

Q: Did the assassinations stop after Fakhrizadeh was killed?

The covert assassination track went largely quiet after the November 2020 killing of Fakhrizadeh, in the sense that no comparable series of street killings of scientists was widely reported in the years immediately following. The campaign against Iran’s nuclear capability did not stop, however. It continued through sabotage of facilities and, decisively, through the open military escalation of 2025, when Israel’s airstrikes killed scientists from the air rather than in traffic. The pattern is best understood not as the campaign ending but as it changing form, moving up the escalation ladder from covert killing to overt military action.

Q: What does the Iran campaign teach India about its own shadow war?

Above all, the central lesson is that tactical success and strategic success are different things. The Iranian campaign killed its targets with professional reliability for fifteen years and still did not achieve its ultimate aim, ending with the very war the covert method was designed to prevent. For India, this is a warning against assuming that a campaign which is successfully eliminating its targets is therefore winning. The Iranian case suggests that targeted killing has a strategic ceiling, that the ceiling is reached sooner than planners expect, and that the hardest and most important decisions concern not how to run the covert campaign but what to do when it reaches its limit. India’s shadow war is an impressive instrument, and the Iranian precedent suggests that instrument’s limits deserve as much attention as its operations.

Q: Could India face the same retaliation Israel faced from Iran?

The Iranian case is a direct caution on exactly this point. For two decades the covert assassination of Iranian scientists did not trigger a full military response, because deniability kept the operation below the threshold that would oblige one. The moment Israel escalated openly in 2025, Iran responded with a major missile barrage and a regional war. India’s campaign in Pakistan currently operates in the same deniable space, and as long as it stays there the risk of full-scale retaliation is constrained. The Iranian precedent warns, however, that this constraint depends entirely on the campaign remaining covert and remaining able to keep pace with the threat. If either condition fails, the Iranian example shows that retaliation, conventional and serious, can arrive quickly once the threshold is crossed.

Q: How is the killing of scientists different from killing terrorists?

The difference is the core moral asymmetry between the Iranian and Indian efforts. India’s targets in Pakistan are men with documented histories of organizing or facilitating terrorist violence, named in charge sheets and on international sanctions lists, against whom a state can point to specific past acts of mass killing. The Iranian targets were scientists and engineers accused of no violence, whose offense was technical competence in service of a program their own government insisted was lawful and civilian. Killing a man for what he has done is easier to defend, legally and morally, than killing a man for what he knows and might help build. Defenders of the scientist killings must rely on a preemptive logic, the future weapon, that the case against a terrorist organizer does not strictly require. This is the single dimension on which the two efforts most sharply diverge, and it is the dimension on which India’s operations are easier to justify.

Q: Could a remote-controlled weapon be used in India’s shadow war?

Nothing in the open record indicates that India has used a remote weapon of the kind reportedly deployed against Fakhrizadeh in 2020, and the Pakistani killings attributed to India have so far relied on the human-operative model of riders and handguns. But the Fakhrizadeh operation proved the remote model is technically achievable, and it solves the single greatest vulnerability of any covert killing program, the human operative who can be caught, interrogated, and used to convert a deniable program into an attributed one. A remote weapon requires significant engineering depth, smuggling capacity, and the ability to emplace and conceal a complex device in hostile territory, so it is not a casual option. Whether India develops or adopts such a capability is unknown, but the Iranian precedent has placed the option on the table, and any serious assessment of where India’s shadow war could evolve has to treat the remote weapon as a live possibility rather than science fiction.

Q: Why is the Iran case relevant to a series about India and Pakistan?

The Iranian effort is the closest functioning real-world model for what India has been doing inside Pakistan, which is why this series examines it in depth. The two share the same signature method, the motorcycle-borne attack using a magnetic charge or a close-range shooting with an escape into dense urban traffic. They share the same posture, widely attributed but never formally claimed. They share the same strategic purpose, degrading an adversary’s capability by attacking the humans who constitute it rather than the infrastructure. And they appear to share the same strategic ceiling, the capacity to delay and harass without delivering a decisive result. The Iranian operation began earlier and has run longer, which means it offers India something close to a forecast: a view of where an effort of this kind leads after fifteen years. For anyone trying to understand India’s shadow war, the Iranian precedent is not a digression but the most instructive case available.

Q: Did the assassination of scientists ever provoke international penalties?

During the covert phase, the killings drew condemnation from Iran and expressions of concern from some governments and human rights organizations, but they did not produce decisive international penalties against the states believed responsible. The deniability of the method was central to that outcome: without formal proof of who carried out a given killing, other governments had limited grounds and limited appetite for sanction. The 2025 escalation was different. When Israel struck Iran openly, the strikes were condemned by the United Nations and many governments, and legal scholars described them as a violation of international law, while a smaller group of states defended them as justified nonproliferation. The contrast between the muted response to fifteen years of covert killing and the loud response to open airstrikes is itself one of the clearest lessons of the case: deniability does not only reduce the risk of war, it also reduces the diplomatic price of the operation.