On a Rome evening in October 1972, a thin, scholarly Palestinian named Wael Zwaiter walked back from a quiet dinner toward his apartment building on the Piazza Annibaliano. He had spent the evening discussing literature. He translated Italian poetry into Arabic and had a gentle reputation among the writers of the city. As he stepped into the dim foyer of his building and pressed the elevator button, two men emerged from the shadows and fired twelve rounds into him at point-blank range. One bullet for each of the eleven Israeli athletes and coaches murdered five weeks earlier in Munich, and one more, by some accounts, for emphasis. The gunmen vanished into a waiting car. No agency claimed the act. No government acknowledged it. A man was simply dead in a stairwell, and the world was meant to draw its own conclusions.

That killing in Rome was the opening move of a campaign that would last more than two decades and reach into a dozen countries. Israel called it Operation Wrath of God, and within the intelligence service it carried the codename Bayonet. It was not a single strike or a brief spasm of revenge. It was a sustained, methodical, deniable programme of state-sponsored assassination conducted against an enemy that lived beyond Israel’s borders, in the cities of friendly democracies, under the protection of governments that would never extradite them. The campaign produced spectacular successes and one catastrophic failure. It killed the right men in Paris and the wrong man in Norway. It generated a body of operational knowledge that intelligence services around the world have studied ever since.

For readers trying to understand why India now hunts terrorists in the streets of Karachi and Lahore, the Rome killing is the place to begin. The argument running through this analysis is direct. Operation Wrath of God was the world’s first sustained state-sponsored assassination campaign against terrorists on foreign soil, and it established the precedent that a functioning democracy can pursue its enemies indefinitely beyond its frontiers. India’s shadow war against terror follows that template with uncanny fidelity: patient preparation, methodical execution, total deniability, and no announced endpoint. To understand the motorcycle assassins of Pakistan’s cities, it helps to first understand the men who waited in a Roman stairwell.

Background and Triggers

The operation cannot be understood without the morning of September 5, 1972, and the catastrophe that unfolded inside the Munich Olympic Village. Eight members of the Palestinian militant faction Black September, an offshoot of Fatah, climbed the perimeter fence of the athletes’ quarters in the early hours and forced their way into the apartments housing the Israeli delegation. Two Israelis were killed in the initial assault. Nine more were taken hostage. What followed was a daylong siege broadcast to a global television audience, a negotiation conducted under the gaze of cameras that had been brought to Bavaria to record sport, and a botched rescue attempt at the Fürstenfeldbruck airfield that ended with all nine remaining hostages dead, along with a West German police officer and five of the attackers. Three of the Black September men survived and were taken into German custody.

The failure at Fürstenfeldbruck deserves to be understood in detail, because it shaped everything that followed. West German authorities had agreed to allow the attackers and their hostages to be transported to the military airfield, ostensibly so the group could board an aircraft and fly out of the country. The plan was to ambush the attackers at the airfield instead. The ambush was undermanned, poorly coordinated, and conducted by police officers without specialized training for a hostage rescue of that complexity. Snipers were too few and badly positioned. Armored vehicles that were meant to support the operation became stuck in traffic and arrived late. When the firefight finally erupted, the attackers turned on the bound and helpless hostages and killed them. For the Israeli leadership watching from a distance, the lesson was not merely that their citizens had died. It was that a friendly European state, given full custody of the situation, had been unable or unwilling to protect them, and that the institutions of Europe could not be relied upon to deliver either rescue or justice.

Black September, the faction responsible, bears its own examination. It was not a fringe band of amateurs. It had emerged as a clandestine offshoot connected to Fatah, the dominant body within the Palestinian movement, and it functioned as a deniable instrument that allowed the broader movement to conduct spectacular violence while maintaining a degree of political distance from it. The Munich attack was its most notorious act, but it was part of a wider campaign of hijackings, bombings, and killings across the early 1970s. The structure mattered to the planners in Jerusalem because it meant that the men who organized Munich were not isolated criminals who could be neutralized by arresting a handful of individuals. They were embedded in a functioning network with financiers, safe houses, forged documents, weapons pipelines, and political cover. To strike at that network meaningfully, the planners concluded, they would have to reach the people who ran it, wherever those people lived.

The shock inside Israel was profound, and it was compounded by a second humiliation. Less than two months after the massacre, a Lufthansa airliner was hijacked, and West Germany released the three surviving Munich attackers in exchange for the passengers. To the Israeli government, the message was unmistakable. The men who had murdered its Olympic team would not face justice through the courts of Europe. They would walk free, give interviews, and be celebrated in parts of the Arab world as heroes. If accountability was going to exist at all, the Israeli state would have to manufacture it itself.

Prime Minister Golda Meir understood the political weight of what she was about to authorize, and she did not want to carry the decision alone. She convened a small and secret group of senior officials, remembered in intelligence histories as Committee X, with Defence Minister Moshe Dayan at its core. The committee’s function was to review intelligence recommendations and decide, individually, which names would be marked for death. This was a deliberate design. By distributing the moral burden across a committee and basing each decision on the dossier presented by the intelligence service, the leadership created a structure that resembled a judicial process while operating entirely outside any court. There were no trials, no defence, no appeal. There was a dossier, a discussion, and a signature.

The committee structure deserves to be examined for what it actually accomplished and what it merely performed. By requiring a small panel of senior figures to consider each dossier and assent to each name, the leadership built something that borrowed the vocabulary of justice without any of its substance. A real trial places the accused in a room, gives him an advocate, tests the evidence against him through adversarial challenge, and submits the verdict to appeal. Committee X did none of this. The accused never knew he had been accused. There was no advocate to argue that the dossier was thin, that the informant was unreliable, or that the man in Rome had been confused with a man of similar name. The evidence was whatever the intelligence service chose to present, assembled by the same institution that would carry out the sentence. What the committee provided was not due process but a distribution of conscience. No single person had to feel that he alone had ordered a man’s death, because the decision had been shared, and a shared decision is psychologically lighter than a solitary one. This is worth dwelling on because it is the precise mechanism by which a lawful democracy reconciles itself to conduct that its own courts would never sanction. The conduct is not made lawful. It is made bearable, and bearability is enough to keep a programme running.

Israel’s first reaction had been conventional and visible. Within two days of the Fürstenfeldbruck disaster, the Israeli Air Force struck around ten Palestinian bases across Syria and Lebanon. Airstrikes, however, satisfied neither the strategic appetite nor the symbolic need of the moment. Bombing a training camp killed foot soldiers and flattened buildings, but it did not reach the planners, the financiers, the logisticians, and the spokesmen who made the network function. The leadership wanted something more discriminating and more frightening: the knowledge, planted in the mind of every Palestinian operative in Europe, that he could be found, that his routine was known, and that the quiet man at the next cafe table might be there to kill him.

The intelligence service was therefore tasked with two distinct jobs. The first was to build a list. Drawing on its own sources, on Palestinian informants, and on cooperation with friendly European agencies, the service assembled a roster of individuals deemed to have planned, financed, facilitated, or otherwise enabled the Munich attack and the broader campaign of Palestinian terrorism. Estimates of the list’s length vary between roughly twenty and thirty-five names, and the full roster has never been published. The second job was operational: to locate each man, study him, and kill him in a manner that could not be traced with certainty back to Jerusalem. Plausible deniability was not a convenience. It was a structural requirement, because the targets lived in Rome, Paris, Nicosia, and Athens, on the soil of states that were nominally friendly to Israel and that would have been obliged to treat an acknowledged foreign assassination as a grave violation of their sovereignty.

This is the first and most important parallel with India’s later campaign. New Delhi faced an almost identical structural problem. The men who planned the 2008 Mumbai attacks and a long series of subsequent atrocities lived openly inside Pakistan, protected by a state that would never hand them over and shielded by an intelligence establishment that treated them as strategic assets. Conventional military action risked nuclear escalation. International legal pressure produced designations and sanctions but no custody. India, like Israel before it, was left with a narrow channel between doing nothing and starting a war, and like Israel it eventually chose to operate inside that channel. The Munich aftermath did not invent the dilemma, but it produced the first fully developed answer to it.

It is worth pausing on how narrow that channel really is, because the narrowness explains why two democracies separated by half a century and an entire continent arrived at the same destination. A state confronting cross-border terrorism has, in theory, a range of options. It can absorb the attacks and rely on defence, hardening its own targets and accepting a steady toll. It can pursue diplomacy and legal process, building cases and demanding extradition. It can apply economic and political pressure, isolating the sponsoring state. It can mount open military action, invading or striking the territory where its enemies shelter. Each of these options, examined honestly, has a disqualifying weakness when the enemy is sheltered by a hostile state that will not cooperate. Pure defence concedes the initiative permanently and never reaches the planners. Legal process depends on a cooperation that the sheltering state refuses to give. Economic pressure is slow and easily absorbed. Open war, against a nuclear-armed adversary in the Indian case or against the complex web of Arab states in the Israeli case, risks a catastrophe out of all proportion to the original injury. What remains, after every broader option has been ruled out, is the narrow channel of covert, deniable, individualized elimination. Neither Israel nor India chose that channel because it was attractive. They chose it because they had closed off every wider one, and the Munich aftermath was the first time a democracy walked that channel from end to end and left a map.

The Opening Strikes in Rome and Paris

The campaign that followed can be read as a chronicle, and reading it operation by operation is the most honest way to understand both its competence and its limits. Each strike had a target, a method, a city, and an outcome, and each one taught the planners something they applied to the next.

The first name on the visible record was Wael Zwaiter, the PLO representative in Italy, killed in Rome on October 16, 1972. Zwaiter was a soft target in the operational sense: he had no bodyguards, no armoured car, and a predictable evening routine. Israel privately maintained that he had been involved in a failed plot against an El Al airliner and had provided logistical support to the network. Palestinian figures, including senior PLO men, insisted he was a man of letters who opposed violence. The dispute over Zwaiter’s actual culpability has never been resolved, and that unresolved quality matters, because it surfaces again and again across the campaign. The method was simple and close: two agents, a stairwell, a pistol, and an immediate extraction to a safe house. The lesson the planners carried forward was that an unguarded man in a European city could be reached with minimal force and minimal exposure, provided his routine had been studied long enough.

That first killing in Rome also established a piece of grim symbolism that the operation would carry forward. Accounts of the Rome shooting record that he was hit roughly a dozen times, and a number of reconstructions describe the count as deliberate, one round for each murdered member of the Israeli Olympic delegation, with an additional shot. Whether or not the symbolism was as precise as later retellings suggest, the campaign clearly understood that its killings were also messages. A body was a statement, addressed not only to the dead man but to every other name on the list and to the network that would learn of the death within hours. This communicative dimension is essential to understanding why the programme worked the way it did. A purely practical campaign would have been content to remove individuals as quietly as possible. This campaign wanted the removals to be noticed, because the fear generated by a noticed killing was itself an operational objective. India’s campaign in Pakistan operates on the same dual logic. The eliminations are never claimed, yet they are never truly hidden either. They occur in public places, they are reported in the Pakistani press, and the pattern across them is legible enough that the intended audience, the wider militant network, receives the message without any official ever having to deliver it.

The intelligence service did not always choose the gun. For its second strike it chose a far more elaborate method, one that revealed how much technical ingenuity the programme could mobilize. The target was Mahmoud Hamshari, the PLO’s representative in France, whom Israel believed to be the senior Black September figure in the country. Rather than confront Hamshari directly, the service ran an agent posing as an Italian journalist who arranged a telephone interview. The pretext drew Hamshari out of his Paris apartment long enough for a technical team to enter and conceal an explosive device inside the base of his desk telephone. On December 8, 1972, the supposed journalist placed the call, waited for Hamshari to identify himself on the line, and transmitted a signal that detonated the charge. Hamshari was not killed instantly. He died of his injuries roughly a month later. The operation demonstrated that the campaign could combine human deception, technical surveillance, and remote killing into a single coordinated act, and it showed a willingness to invest weeks of preparation in a single name.

The third strike returned to the bomb but moved the theatre to the Mediterranean. Hussein al-Bashir, sometimes recorded as Hussein Abad al-Chir, was the Fatah representative in Cyprus and, in the Israeli assessment, an important conduit between the Palestinian network and its Soviet-bloc contacts. On the night of January 24, 1973, al-Bashir switched off the light in his room at the Olympic Hotel in Nicosia. Moments later a device that had been planted beneath his bed was detonated by remote signal, killing him and destroying the room. The Cyprus killing extended the hunt’s geographic reach and confirmed a pattern: the planners would follow a target to whatever city he occupied, and a hotel room was no safer than an apartment.

Hamshari’s killing in particular rewards a closer look, because it shows how much the campaign was willing to spend, in time and ingenuity, to kill a single man cleanly. Installing an explosive charge inside the body of a telephone is not a simple act. It requires entry into the target’s home without his knowledge, which requires first removing him from that home for a reliable interval, which is what the false journalist accomplished. It requires a device small enough to be concealed, reliable enough to function on command, and powerful enough to be lethal at close range without destroying the building. It requires a triggering mechanism keyed to the target’s own voice on the line, so that the charge fires when the right man, and only the right man, has answered. And it requires a team patient enough to wait for the interview to be arranged, for the call to connect, and for the identification to be confirmed before the signal is sent. The whole sequence is a small masterpiece of operational planning, and it was assembled to kill one PLO representative in one Paris apartment. The willingness to invest at that level is the effort’s defining characteristic, and it is the characteristic India’s planners most clearly inherited. The reported eliminations in Pakistan are not the work of opportunists. They are the product of weeks of surveillance, of mapped routines and identified vulnerabilities, of a patience that treats a single target as worth a month of preparation.

Across these opening strikes, a recognizable operational signature emerged. Targets were studied for weeks. Their routines, their residences, and their movements were mapped before any agent acted. The methods were varied deliberately so that no single countermeasure could defeat the campaign. The teams operated in small cells, executed quickly, and extracted immediately. Above all, the strikes were spaced and patient. There was no rush to clear the list. The planners behaved as though they had unlimited time, because in a sense they did.

The parallel with India’s campaign in Pakistan is precise enough to be uncomfortable. The eliminations attributed to India in Karachi, Lahore, Rawalpindi, and Sialkot share that same signature: extended surveillance of a target’s daily pattern, a strike delivered at a predictable location such as a mosque at prayer time or a residence at a fixed hour, a small team, and an immediate disappearance into urban congestion. The method in Pakistan has tended toward the motorcycle-borne shooter rather than the telephone bomb, but the underlying doctrine is the same one the intelligence service refined in Rome, Paris, and Nicosia. The Israeli campaign proved that a database of routines, patiently built, is a more powerful weapon than any single bomb. India’s planners absorbed that proof and built their own database.

Operation Spring of Youth and the Paris Killings

By the spring of 1973 the operation had demonstrated that it could reach soft targets in open cities. It had not yet demonstrated that it could reach hardened targets in hostile territory, and the planners were determined to prove that it could.

First, however, the campaign returned to Paris. On April 6, 1973, Basil al-Kubaisi, a law professor at the American University of Beirut whom Israel suspected of arranging arms logistics for Palestinian operations, was shot roughly a dozen times by two agents as he walked home from dinner. The al-Kubaisi killing reinforced the Roman model. An academic with a public life and a predictable schedule could be reached on a city street, and the repetition of the close-range pistol method showed that the planners were content to reuse a technique that worked rather than reinvent one for every name.

Three days later came the operation that announced, more loudly than any other single act, that the programme would not be confined to soft targets. Several men on the list lived in heavily guarded apartment blocks in Beirut, surrounded by armed Palestinian fighters, in a city Israel could not enter through ordinary covert means. The answer was Operation Spring of Youth, a military raid folded into the larger assassination programme. On the night of April 9 into April 10, 1973, commandos from the elite Sayeret Matkal unit, supported by naval commandos and paratroopers, came ashore on the Lebanese coast in inflatable boats launched from missile craft offshore. Mossad agents already in Beirut met them with rental cars. Some of the raiders were disguised as civilians, and a number were dressed as women in order to move through the streets without alarming sentries. The teams reached their target buildings, fought through the guards, and killed three senior Palestinian figures: Muhammad Youssef al-Najjar, known as Abu Youssef, an operations leader linked to Black September; Kamal Adwan, a senior PLO operations chief; and Kamal Nasser, a member of the PLO executive committee and the movement’s spokesman. A future Israeli prime minister, Ehud Barak, was among the raiders. The teams withdrew to the coast and were extracted by sea.

Operation Spring of Youth changed the strategic meaning of the campaign. It proved that no sanctuary was absolute. A target who believed that bodyguards and a fortified building placed him beyond reach was simply a target who had not yet been reached. The raid also blurred the line between covert assassination and conventional military action, a blurring that intelligence services have exploited ever since. India’s campaign has produced its own version of this lesson. The reported reach of the eliminations into Pakistan-occupied Kashmir, into locations close to the Line of Control, and into the supposedly secure interiors of Pakistan’s largest cities carries the same message that Beirut carried in 1973: protection is a comfort, not a guarantee.

The detail of how the raid was conducted is worth recording, because it reveals the hunt’s range. The teams did not improvise. They rehearsed. Naval craft carried the commandos to a point off the Lebanese coast, from which inflatable boats brought them ashore in darkness. Intelligence officers who had been operating inside Beirut for weeks had already rented the cars that would carry the raiders to their targets, scouted the routes, and identified the buildings. The disguises, including the decision to dress some raiders as women and others as ordinary civilians, were chosen so that armed groups of men moving through a tense city at night would not trigger an alarm before the teams reached their objectives. After the killings, the same cars carried the raiders back to the shoreline, and the same boats carried them back to the ships. Every phase, insertion, movement, execution, and extraction, had been planned as a continuous sequence. This is the difference between a killing and an operation. A killing is an event. An operation is an architecture, and Spring of Youth demonstrated that the campaign could build an architecture complex enough to reach into the heart of a hostile capital and withdraw before the city had fully understood what had happened.

The strategic ambition on display in Beirut is precisely what makes the Israeli campaign a template rather than a mere series of murders. A state that can only kill the unguarded has a limited instrument, useful against soft targets and useless against anyone who takes precautions. A state that can also, when it chooses, mount a full military raid against a fortified target has an instrument with no upper limit. The enemy cannot find safety by adding bodyguards, because bodyguards can be fought through. He cannot find safety by retreating into a friendly capital, because the capital can be entered. The psychological effect of that unlimited reach is larger than the sum of the individual killings. It tells every potential target that there is no posture of safety available to him, and a man who believes he can never be safe is a man whose effectiveness has already been degraded, whether or not he is ever actually attacked.

That spring of 1973 was not yet finished. On April 11, the day after the Beirut raid, the effort struck again in the Mediterranean. Zaiad Muchasi, who had stepped into the Cyprus role left vacant by the killing of al-Bashir, was killed by an explosive device in his Athens apartment. The Muchasi killing carried a lesson that the strategic debate would later seize upon. Eliminating al-Bashir had not closed the position. It had simply created a vacancy, and the vacancy had been filled. The programme was discovering, strike by strike, that an organization is not a fixed set of individuals but a structure that regenerates.

The Muchasi case is worth holding in mind, because it is the campaign’s own internal evidence against the strategic theory that justified it. The planners had killed al-Bashir on the premise that removing him would degrade the network’s capability in Cyprus. The network responded by installing Muchasi, and the operation then had to expend a second operation, with all its attendant risk and cost, to kill the replacement. If the logic of the campaign were sound, this should have given the planners pause. Each killing that produces a successor is not a step toward victory. It is a step onto a treadmill, an exchange of operational risk for the temporary removal of a function that the enemy immediately restores. The campaign documented this dynamic in its own files and continued anyway, which tells us something important about how such programmes behave: the institutional momentum of a hunt, once begun, tends to override the evidence the hunt itself generates.

The last major strike of this intense phase came on June 28, 1973, when Mohamed Boudia, an Algerian-born operative associated with Palestinian operations in Europe and a figure of considerable theatrical and operational cunning, was killed by a car bomb in Paris. Boudia had been notoriously difficult to track, a man of disguises and false identities, and his death marked the end of the programme’s most concentrated period of activity. In roughly nine months the programme had killed a series of significant figures across Italy, France, Cyprus, Greece, and Lebanon. It had used pistols, telephone bombs, bed bombs, car bombs, and a full amphibious commando raid. It had built, in less than a year, a working model of how a state could hunt a dispersed enemy across a continent.

That model is the inheritance India later drew upon. The variety of methods, the willingness to invest months in a single target, the readiness to escalate from a lone gunman to a military raid when the target’s protection demanded it, and above all the patience: all of it reappears in the pattern of eliminations attributed to India across Pakistan. The Israeli campaign did not merely kill men. It generated a doctrine, and doctrine travels.

Before turning to the programme’s great failure, it is worth drawing the operational lessons of this first phase together into a single picture, because the picture is the actual artifact the hunt produced. Across roughly nine months, the programme had demonstrated a set of repeatable principles. The first principle was target study: no strike was attempted until the target’s routine had been mapped in enough detail to predict where he would be and when. The second principle was method variation: pistols, telephone bombs, bed bombs, car bombs, and commando raids were used in deliberate rotation so that no single defensive measure could protect the network as a whole. The third principle was small-cell execution: the actual killing was carried out by compact teams that minimized exposure. The fourth principle was immediate extraction: the team’s escape was planned with as much care as the strike, because an operation that kills the target but loses the team is a failure. The fifth principle was patience: the campaign accepted that the list would be cleared over years, not weeks, and refused to be hurried into sloppy work by the desire for speed. The sixth principle was deniability: every strike was designed so that no acknowledged connection to the state could be proven, even when the broader pattern made the responsible party obvious.

Set those six principles side by side with the documented signature of India’s campaign in Pakistan and the overlap is nearly total. The eliminations attributed to India are preceded by extended surveillance. They vary in method, though they favour the motorcycle-borne shooter. They are carried out by small teams. They are followed by clean disappearances into urban traffic. They are spaced across years rather than compressed into a single burst. And they are never, under any circumstances, officially acknowledged. The Israeli campaign of the early 1970s is not a vague historical inspiration for the Indian programme. It is something far more specific: a fully worked operational model, refined under real conditions, that India’s planners could study and apply. The chronicle of Rome, Paris, Nicosia, Athens, and Beirut is, in effect, the first edition of a manual that is still in use.

The Lillehammer Disaster

Every honest account of Operation Wrath of God arrives, sooner or later, at a small town in Norway, because the effort’s defining failure is as instructive as any of its successes, and arguably more so.

By the summer of 1973 the planners were fixed on a single name above all others: Ali Hassan Salameh, the Fatah and Black September operations figure whom Israel held to be a principal architect of the Munich massacre. Salameh had acquired a near-mythic status inside the intelligence service. He was elusive, well-protected, and personally important to the campaign’s sense of completion, and he was given a nickname that captured both the fascination and the frustration he generated: the Red Prince.

In July 1973 the service believed it had traced Salameh to Lillehammer, a quiet Norwegian ski town far from the cities where the operation had so far operated. A team of around a dozen operatives assembled. What happened next has been disputed in its details, but the essential outline is clear and damning. The team appears to have lost track of its actual target and, using a method of narrowing down a suspect’s likely location across a search area, fixed instead on a man they came to believe was connected to the Palestinian network. The man they identified was not Ali Hassan Salameh. He was Ahmed Bouchikhi, a Moroccan-born waiter who worked in the town and helped maintain its swimming pool. He had a Norwegian wife who was seven months pregnant.

On the night of July 21, 1973, as Bouchikhi and his wife left a cinema and walked toward a bus stop, two operatives shot him to death in front of her. He had no connection to Black September, to the Munich attack, or to any act of terrorism. He was an innocent man killed by a state intelligence service that had convinced itself, on inadequate evidence, that he was someone else.

The Lillehammer affair did not end with the killing. Norwegian police, investigating an assassination on their own soil, moved with a competence the planners had not anticipated. Within days they had arrested members of the team. The arrests, and the interrogations that followed, exposed far more than a single botched operation. They unravelled a portion of the service’s network of safe houses, agents, and infrastructure across Europe, because a captured operative is a thread, and pulling a thread unravels a fabric. Six people were tried in Norway in connection with the killing, and five were convicted, though they served relatively short terms and were later released and returned to Israel. The reputational damage, however, could not be released on parole. A campaign built on plausible deniability had produced a courtroom, a verdict, and a dead innocent whose name the world now knew.

The mechanics of how the operation came apart are themselves a lesson in the fragility of covert work. The team in Lillehammer had not maintained the operational discipline that the earlier strikes displayed. After the killing, members of the team were traced through a rented vehicle, and the vehicle led investigators to others. Once the first operatives were in custody, the Norwegian authorities possessed something no intelligence service can afford to surrender: living witnesses, in a cell, subject to questioning, with every incentive to explain themselves. The information that emerged compromised safe houses and contacts that had nothing to do with Lillehammer and everything to do with the wider European campaign. A programme that had spent a year building an invisible infrastructure watched a meaningful part of that infrastructure become visible because of a single operation conducted with insufficient care. The episode demonstrates a principle that every covert campaign would prefer to ignore: deniability is not a permanent property of an operation. It is a temporary condition that survives only as long as nothing goes wrong, and the probability that nothing will ever go wrong, across a campaign of many operations and many years, is not high.

Norway’s courtroom also did something the operation’s planners could not have wanted. It converted an abstraction into a person. Throughout the programme, the targets had been names on a list, figures in dossiers, men whose deaths could be discussed in the language of operations and objectives. Ahmed Bouchikhi could not be discussed in that language, because he had no dossier, no role, and no connection to the network. He was a waiter. He cleaned a swimming pool. He had a wife who was carrying their child. The trial placed his ordinariness on the public record, and in doing so it stripped the campaign of the protective abstraction that had allowed it to operate. For a moment, the world was not looking at an intelligence operation. It was looking at the body of an innocent man and at the agents of a state standing trial for putting it there.

The response in Jerusalem was decisive. Golda Meir, facing intense international condemnation and the exposure of the service’s European apparatus, ordered the suspension of the assassination programme. The hunt that had moved with such confidence from Rome to Beirut was halted not by its enemies but by its own catastrophic error.

The Lillehammer disaster deserves close attention because it isolates the single most dangerous vulnerability in any targeted-killing programme, and that vulnerability is identification. Every other element of the hunt can function perfectly. The surveillance can be patient, the team can be disciplined, the extraction can be clean, and none of it matters if the wrong man has been placed under the crosshairs. Intelligence services do not kill people. They kill their assessment of people, and when the assessment is wrong, the gap is filled by the corpse of someone who did nothing.

This is the warning that runs directly into the analysis of India’s shadow war, and it is the reason a serious account cannot present the Israeli campaign as a simple success story to be admired and copied. India’s eliminations in Pakistan have repeatedly raised the same question that Lillehammer raised. When a man is shot dead by unidentified attackers in Karachi or Sialkot, the public is asked to accept, on the basis of anonymous briefings and circumstantial pattern, that the dead man was a terrorist who deserved his fate. Sometimes the evidence for that claim is substantial. Sometimes, as Pakistan’s own counter-narrative insists, the dead man’s connection to militancy is contested or thin. Lillehammer is the permanent reminder that a state operating without courts, without disclosure, and without independent review is a state that has removed every mechanism by which an identification error could be caught before it becomes fatal. The Israeli campaign did not merely teach India how to hunt. It also left, in a Norwegian courtroom, the clearest available demonstration of how the hunt can go terribly wrong.

It is worth being precise about why the identification problem is structural rather than incidental. A targeted-killing programme that operates covertly cannot use the ordinary safeguards against error. In a criminal justice system, an identification can be tested before the irreversible step is taken. The accused can be brought into a courtroom and confronted with witnesses. Documents can be examined. An advocate can argue that the man in the dock is not the man the evidence describes. Every one of these safeguards exists precisely because human institutions know that identification is fallible, that informants lie or are mistaken, that names recur, that faces resemble one another, and that an intelligence assessment is a hypothesis rather than a fact. A covert killing programme discards all of these safeguards by design, because the entire point of the programme is to act without the visibility that a courtroom would require. The result is that the only check on the identification is the confidence of the officers who made it, and confidence is not a check. The Lillehammer team was confident. It was also wrong. The structural lesson is that a programme which removes every external test of its identifications is a programme that has guaranteed, as a matter of statistics rather than bad luck, that some of its identifications will be wrong, and that the errors will only become visible after a body is on the ground. India’s planners, operating with even less acknowledged oversight than the Israeli system eventually developed, are exposed to exactly this structural risk, and no amount of operational skill can eliminate it, because the risk lives in the identification, not in the execution.

The Long Hunt and the Final Reckoning

A campaign suspended is not a campaign abandoned, and the men who ran Operation Wrath of God did not consider the Red Prince a closed file.

Ali Hassan Salameh survived not only the Lillehammer fiasco but a string of further attempts on his life over the following years, including reported plots in other European cities and a failed effort in Spain. He had become, by the late 1970s, both the campaign’s great unfinished business and the proof that even a determined intelligence service could be outlasted by a careful man.

The years between Lillehammer and the final reckoning are instructive precisely because they were years of failure. The campaign did not glide smoothly toward its conclusion. It stumbled. Reports describe an aborted effort against Salameh connected to a planned operation in Britain, and a separate attempt in continental Europe that came apart before it could be completed. In one episode, a team that had tracked Salameh to a location in Spain was forced to abandon the operation after an encounter with an armed guard, and the operatives withdrew without reaching their target. Each of these failures carried a cost. Every aborted attempt risked exposing operatives, burning surveillance assets, and alerting the target that he remained hunted. Salameh, for his part, had learned the lessons of the effort as thoroughly as the campaign had learned its trade. He varied his movements, maintained protection, and understood that he was the most wanted name on a list that did not expire. For roughly six years, a careful man imposed delay and cost on one of the world’s most capable intelligence services, and the delay is itself a finding. It demonstrates that a targeted-killing campaign, however skilled, does not enjoy the certainty its reputation suggests. It can be resisted. It can be slowed. A sufficiently disciplined target can convert his own life into a prolonged and expensive operation for his hunters.

The reckoning came in January 1979 in Beirut. The programme was reactivated for what amounted to a final mission. The service had studied Salameh’s movements and identified a route he travelled regularly through the city. A vehicle packed with explosives was positioned along that route, and as Salameh’s convoy passed, the charge was detonated by remote signal. He was killed, along with several bystanders and members of his protective detail. The killing of Salameh, more than six years after the Munich massacre that set the operation in motion, closed the chapter that the planners cared about most. It also illustrated a defining quality of the entire enterprise. The gap between the crime and the final act of retribution was measured not in weeks or months but in years, and the campaign had simply absorbed that span without losing focus.

Even Salameh’s death did not formally end the broader programme. Killings attributed to the Israeli service continued against Palestinian figures through the 1980s and into the 1990s. The assassination of Atef Bseiso in Paris in 1992, two decades after the Rome killing of Zwaiter, is often cited as a late echo of the same campaign. Whether or not every such killing belonged strictly to the original operation, the cumulative effect was to establish a simple and chilling principle: the Israeli state did not forget, and it did not impose a statute of limitations on its own memory.

This quality of permanence is the parallel that should trouble and clarify any reader thinking about India’s campaign. The eliminations attributed to India did not begin and end in a single year. They have continued across a span of years, accelerating rather than fading, reaching from the established theatres of Karachi and Lahore outward toward new geographies. The Israeli campaign demonstrated, decades earlier, that a targeted-killing programme run by a determined democracy has no natural expiry date. It runs until the political will that sustains it disappears, and political will of that kind tends to be durable. India’s planners did not need to invent the concept of a permanent hunt. They inherited it.

The long arc from Rome in 1972 to Beirut in 1979 to Paris in 1992 also reframes what the programme actually was. It was not an operation in the conventional sense of a discrete mission with a start and a finish. It was closer to a standing institution, a permanent capability that could be activated, suspended, and reactivated as circumstances allowed. That institutional character is precisely what makes it the template rather than merely a precedent. A precedent is a single example. A template is a reusable design, and Operation Wrath of God was a reusable design for how a state hunts its enemies across borders without ever admitting that it is doing so.

Key Figures

Operation Wrath of God was a system, but systems are built and run by individuals, and four figures in particular illuminate how the campaign functioned and what it cost.

Golda Meir

The prime minister who authorized the hunt carried a political reputation built on toughness, and the decision to pursue the Munich planners is inseparable from her name. Yet the most revealing thing about Meir’s role is not the toughness but the structure she built around the decision. By creating Committee X and insisting that each name be reviewed and approved individually rather than handed to the intelligence service as a blanket licence, she constructed a process that distributed responsibility and imposed at least the form of deliberation. Her later decision to suspend the campaign after Lillehammer is equally revealing. It showed that the political authority behind the programme retained the capacity to halt it when the costs became intolerable. The contrast with India is instructive here. India’s campaign operates without any acknowledged authorizing structure at all, which means it also operates without any visible mechanism for being stopped.

Michael Harari

The operational direction of much of the effort is associated with Michael Harari, a senior intelligence officer who oversaw the European teams. Harari embodied the professional ideal the campaign aspired to: meticulous, patient, willing to invest months in preparation, and focused on extraction and deniability as much as on the kill itself. He was also connected to the chain of decisions that led to the Lillehammer disaster, which made him a figure of both competence and controversy. Harari’s career is a reminder that a targeted-killing programme is only ever as reliable as the judgment of the officers who run its cells, and that judgment, however professional, is fallible.

Ali Hassan Salameh

Salameh, the Red Prince, was the operation’s defining adversary, and his importance was as much symbolic as operational. As the most wanted name on the list, Salameh organized the programme’s sense of purpose. The pursuit of him produced both its worst failure, at Lillehammer, where a team hunting him killed an innocent man, and one of its most determined successes, the 1979 Beirut car bomb. Salameh’s long survival demonstrated that a careful, well-protected target could impose enormous costs on the hunters, and his eventual death demonstrated that sufficient patience could still close even the hardest file.

Zvi Zamir

The director of the intelligence service during the programme’s most active years, Zvi Zamir, had been present near the Munich airfield during the failed rescue and carried that experience into the operation that followed. Zamir’s recorded reflections on the campaign, including his comment that a wrong identification was a mistake rather than a failure, capture the moral posture of the enterprise: a willingness to accept error as an inherent cost of the work rather than as a disqualifying objection to it. That posture is the quiet, central assumption of every targeted-killing programme, including India’s, and it deserves to be named plainly rather than left implicit.

The distinction Zamir drew between a mistake and a failure repays careful thought, because it is the hinge on which the entire moral architecture of the hunt turns. To call the killing of an innocent man a failure would be to say that the operation should not have happened, that the system that produced it was broken, and that the programme itself stood condemned by the result. To call it a mistake is to say something far less damaging: that the operation was sound, that the system was working as intended, and that the death of the wrong man was an unfortunate outcome of a process that remains legitimate. The choice of word is not casual. It is the mechanism by which a programme survives its own worst outcomes. If every error were treated as a failure, the campaign would have to stop after the first one. By treating error as a mistake, an expected and absorbed cost, the effort grants itself permission to continue. Every covert killing programme in history has needed some version of this distinction, and every reader assessing such a programme should notice when it is being deployed, because it is the precise point at which an institution decides that the deaths of innocents will not be allowed to halt its work.

Consequences and Impact

Operation Wrath of God produced consequences that extend well beyond the body count, and they fall into several distinct categories.

The most immediate consequence was operational disruption of the Palestinian network in Europe. A series of representatives, logisticians, and organizers were removed, and the survivors were forced to alter their behaviour, restrict their movements, and divert energy from operations toward personal security. Whether this disruption translated into a measurable reduction in terrorism is a question the analytical debate must address, but the disruption itself is not in doubt. An organization that must spend its attention guarding its own people is an organization with less attention to spend on attacking others.

A second consequence was the regeneration problem the campaign itself exposed. The replacement of the assassinated Cyprus figure by a successor, who was then also assassinated, demonstrated that killing an individual does not kill the function he performed. Organizations adapt. They promote, they reassign, and in some cases the successor proves harder and more capable than the man he replaced. The effort therefore generated, almost as a by-product, one of the strongest arguments against the strategic value of targeted killing: that it can decapitate without dismantling, and that decapitation invites a more dangerous head to grow.

A third consequence was reputational and legal. The Lillehammer affair produced a courtroom verdict, the exposure of an intelligence network, and a lasting stain that no operational success could fully erase. It demonstrated that a deniable campaign is only deniable until it is not, and that a single error can convert a hidden programme into an international scandal. The campaign also generated a permanent debate about legality, because the killings were carried out on the soil of states that had not consented, against individuals who had been convicted by no court.

Cultural and doctrinal effects made up a fourth consequence, and it is the one most relevant here. Operation Wrath of God entered the imagination of states and intelligence services as a model. It was dramatized in film, studied in professional literature, and absorbed into the doctrine of how a democracy might pursue terrorists beyond its borders. The operation’s cinematic afterlife, including its treatment in Spielberg’s Munich, kept it alive in public consciousness for decades, and the comparison between that sober, anguished cinematic treatment and the more celebratory tone of India’s own cultural depictions of covert operations is itself revealing about how the two democracies process state violence. The most important consequence of the operation may be the simplest to state: it made the unthinkable into the precedented, and a precedent, once set, is available to be followed.

A fifth consequence is harder to measure but should not be omitted, and it concerns the internal effect on the institution that conducted the campaign. An intelligence service that has hunted and killed across a continent for years is not the same institution it was before. It has acquired capabilities, but it has also acquired habits, assumptions, and a sense of what is permissible. The accumulated experience of Operation Wrath of God shaped the service’s self-understanding and its doctrine for decades afterward, contributing to a broader institutional posture toward targeted killing that successive generations of officers inherited. This matters for the comparison with India because it suggests that a campaign of this kind is never fully reversible at the level of the institution. A government can suspend a programme, as Golda Meir did. It cannot un-teach the skills, un-form the habits, or un-make the assumption, now lodged inside the institution, that hunting enemies abroad is a normal part of the work. India’s intelligence establishment, having developed and exercised its own elimination capability, has undergone the same kind of internal transformation, and that transformation will outlast any particular phase of the programme.

A sixth consequence concerns the relationship between the campaign and the states on whose soil it operated. The killings took place in Italy, France, Cyprus, Greece, Norway, and Lebanon, among others, and each strike was a violation of the host state’s sovereignty and a problem for its security services. Some European agencies cooperated, formally or informally, with the broader intelligence-sharing arrangements of the era, and the question of how much they knew and when has occupied historians since. The campaign therefore generated not only dead targets but a web of diplomatic and inter-agency consequences, some of which strained relationships and some of which were quietly managed. India’s campaign has produced its own version of this web. The eliminations attributed to India, and the parallel allegations connected to incidents on Western soil, have generated diplomatic friction with several states and a sustained international argument about the acceptability of the practice. The Israeli campaign was the first demonstration that a cross-border killing programme generates a diplomatic cost that must be continuously managed, and that the management of that cost is itself a permanent feature of running such a programme.

Analytical Debate

The central analytical question about Operation Wrath of God is deceptively simple and genuinely unresolved. Did the hunt achieve its strategic objective, or did it merely satisfy a need for revenge while changing nothing of consequence? Serious analysts have lined up on both sides, and an honest treatment must give each its full weight.

The case that the campaign worked rests on several pillars. The first is deterrence through fear. Defenders of the programme argue that the visible, unpredictable killing of Palestinian operatives across Europe imposed a psychological cost on the entire network. An operative who knows he might be found, whose telephone might be a bomb and whose evening walk might be his last, is an operative who is distracted, cautious, and less effective. The second pillar is disruption. The removal of experienced figures, defenders argue, degraded the network’s capacity in concrete ways, even if successors eventually emerged, because institutional knowledge and personal relationships are not instantly replaceable. The third pillar is a counterfactual claim: that the relative decline of Black September as a distinct operational entity in the years after the effort suggests, even if it does not prove, that sustained pressure contributed to the group’s diminishment.

An equally substantial case holds that the campaign failed, or at best achieved little of strategic value. Critics point first to the regeneration problem the operation itself revealed. If killing a figure simply produces a successor, and if the successor is sometimes more radical, then the campaign is running on a treadmill, expending enormous risk and resource to stay in roughly the same place. Critics point second to the retaliation problem. The Palestinian network did not absorb the programme passively. It struck back, continued its own attacks, and even planned operations against Israeli leadership, which suggests that the campaign provoked as much as it deterred. Critics point third to the cost of error. Lillehammer killed an innocent man, exposed an intelligence network, and generated a lasting reputational wound, and these costs must be weighed against whatever the hunt achieved. Critics point fourth to the deeper question of root causes. Assassination addresses individuals, not the political grievances, recruitment pipelines, and ideological currents that produce a steady supply of new individuals. A campaign that kills planners faster than it can kill the conditions that create planners is, on this reading, a tactic mistaken for a strategy.

The most intellectually honest position is that the evidence does not permit a clean verdict, and that the inability to reach one is itself the most important finding. Because the campaign was covert, because its full target list was never published, and because the counterfactual world in which it did not happen cannot be observed, the question of its effectiveness is structurally unanswerable with confidence. This matters enormously for the analysis of India’s shadow war, because India’s planners are conducting their campaign in the same fog. They cannot prove that the eliminations in Pakistan have prevented attacks, because the attacks that did not happen cannot be counted. They are operating, as the Israeli planners operated, on a strategic intuition that pressure works, an intuition that is plausible, that is widely shared, and that has never been and perhaps never can be verified. A campaign justified by an unfalsifiable claim is a campaign that will always be able to declare itself a success and will never be able to prove it.

The unfalsifiability problem deserves to be stated even more sharply, because it is the single most important analytical point in any honest assessment of targeted killing. Consider the two possible outcomes after a campaign of eliminations. If terrorist attacks decline, the effort’s defenders will attribute the decline to the campaign. If terrorist attacks continue or increase, the operation’s defenders will argue that the attacks would have been worse without the campaign. Notice that both outcomes are read as evidence that the programme worked. There is no outcome that the campaign’s logic treats as evidence of failure, because the relevant comparison, the world in which the hunt did not happen, is permanently invisible. A claim that cannot be contradicted by any possible observation is not a strategic assessment. It is a faith. This does not mean the campaign was useless. It may well have imposed real costs on its enemies. It means only that the people running it, then and now, do not actually know whether it achieved its strategic purpose, and that their confidence, however sincere, rests on an intuition that the structure of a covert campaign makes it impossible to test. Any reader assessing India’s shadow war should hold this point firmly. The programme’s success is asserted. It is not, and probably cannot be, demonstrated.

There is a second analytical lens worth applying, which is the distinction between a tactic and a strategy. A tactic is a method of achieving a specific, limited objective. A strategy is a coherent plan for reaching a desired end state. Targeted killing is unambiguously a tactic: it can remove a specific individual, and it can do so with precision. Whether it is also a strategy depends on whether the removal of individuals adds up, over time, to a desired end state, such as the permanent diminishment of the enemy network or the deterrence of future attacks. The regeneration problem suggests that the addition does not reliably work, because the network refills the vacancies the effort creates. If that is correct, then a campaign of targeted killing is a tactic being asked to do the work of a strategy, and a tactic mistaken for a strategy is a recipe for indefinite effort without a defined conclusion. This is precisely the condition that the Israeli campaign settled into, a standing capability with no endpoint, and it is the condition that India’s shadow war appears to have entered as well. The honest analytical conclusion is not that targeted killing is worthless, but that it is being asked to deliver an outcome, the strategic defeat of an adversary, that the tactic by itself is structurally unable to deliver.

There is a further dimension of the debate that deserves its own attention, and it concerns accountability rather than effectiveness. Israel’s campaign, for all its secrecy, operated with a degree of structure that India’s does not. There was an authorizing committee. There was a prime minister who took responsibility for activating the programme and who took responsibility for suspending it. In later decades the Israeli system developed a measure of legal scrutiny over targeted killing, including judicial consideration of the practice. India’s campaign, by contrast, is conducted in a posture of total denial. The government does not acknowledge the eliminations, does not describe an authorizing process, and does not submit the practice to any visible review. Some analysts argue that this total deniability is an operational strength, because what is never admitted is harder to constrain and harder to use against the state diplomatically. Others argue that it is a democratic danger, because a programme that is never acknowledged is also a programme that can never be debated, corrected, or stopped by the public in whose name it is carried out. This is the disagreement that the direct comparison between the Indian and Israeli programmes turns on, and it is not a disagreement that a single article can resolve. It can only be named clearly so that readers can weigh it for themselves.

Why It Still Matters

A campaign that began with a killing in a Roman stairwell in 1972 might seem like distant history, of interest to students of the Cold War and little more. The argument of this analysis is that the opposite is true, and that Operation Wrath of God matters now precisely because it is being repeated, in a different country, against a different enemy, on a different continent, with a fidelity that is not coincidental.

Consider what India’s campaign and the Israeli campaign share, point by point. Both were the response of a democracy to a terrorist atrocity that exposed the inadequacy of conventional and legal remedies. Both confronted an enemy who lived openly on the soil of a state that would not extradite him. Both chose a strategy of patient, methodical, deniable elimination over either inaction or open war. Both built an operational signature based on extended surveillance of a target’s routine and a strike delivered at a predictable location. Both varied their methods to defeat countermeasures. Both demonstrated that no sanctuary, however fortified, is truly beyond reach. Both accepted error as an inherent cost of the work. Both treated the campaign as a standing capability with no announced endpoint rather than as a discrete mission. The list of shared features is long enough that the relationship cannot be dismissed as parallel evolution. India’s planners studied the Israeli model, and the intelligence partnership between the two states provided a channel through which doctrine, technology, and operational confidence could move.

The deeper inheritance, however, is conceptual rather than technical. What Operation Wrath of God really bequeathed was a permission structure, an answer to the question of whether a state that considers itself lawful and democratic can hunt and kill its enemies abroad without destroying its own character. Israel’s campaign answered that question in the affirmative, and the world, for the most part, accepted the answer. The accumulated doctrine of Israeli targeted killing, sometimes summarized through the lens of the Rise and Kill First analysis of the practice, became an available intellectual resource for any state facing a similar dilemma. India’s shadow war is, in the most literal sense, an exercise of the permission that the Israeli campaign established.

There is a precedent older than Wrath of God that should be acknowledged here, because the Israeli campaign did not emerge from nothing. The capture of Adolf Eichmann in Argentina more than a decade earlier had already demonstrated that a state could send agents onto foreign soil to seize a wanted man, in defiance of extradition norms and without the host state’s consent. That earlier operation established the extraterritorial principle in its capture form: the idea that a wanted individual abroad was still within the reach of a determined state. Wrath of God took the same extraterritorial principle and converted it from capture into killing. The progression is important. The earlier operation proved that sovereignty could be penetrated to seize a man and bring him to trial. The later campaign proved that sovereignty could be penetrated to kill a man with no trial at all. India’s shadow war stands at the end of that progression, inheriting both the extraterritorial principle and its lethal application.

But the inheritance is double-edged, and a reader who takes only the operational lessons and ignores the cautionary ones has misread the history. Operation Wrath of God also bequeathed Lillehammer. It bequeathed the regeneration problem, the spectacle of an assassinated figure simply being replaced. It bequeathed the unfalsifiability problem, the inability of a covert campaign to ever prove that it worked. It bequeathed the retaliation problem, the evidence that killing can provoke as much as it deters. It bequeathed the accountability problem, the danger of a programme that operates beyond the reach of courts and public debate. India has inherited all of this, not merely the parts that flatter the operation. The motorcycle assassin in a Pakistani city is operating inside a tradition that includes both the clean strike in Paris and the dead waiter in Norway, and the tradition does not allow its inheritors to keep one without risking the other.

It is also worth noting where the two campaigns diverge, because the divergences are as instructive as the parallels. Israel’s campaign, over its long life, developed at least some acknowledged structure: an authorizing committee at the outset, and in later decades a measure of judicial consideration of the targeted-killing practice. Israel’s posture became one of semi-acknowledgment, neither full confession nor total denial. India’s campaign, by contrast, has been conducted in a posture of complete denial, with no acknowledged authorizing process and no visible review of any kind. This is not a small difference. A semi-acknowledged programme can at least be argued about in public, constrained by courts, and debated by a parliament. A wholly denied programme exists, by design, outside the reach of all three. Whether India’s deeper deniability is a strength or a danger is genuinely contested, and the direct comparison of the two doctrines is where that contest is best examined. The point for this analysis is narrower: India did not simply copy the Israeli template. It adopted the operational core while stripping away even the limited accountability scaffolding that the Israeli system eventually built, which means India’s campaign carries the template’s risks in a more concentrated form.

This is why the events of the early 1970s remain urgent. They are not a closed chapter of history. They are the first complete run of a programme that is now running again. The Israeli effort is the rough draft, and India’s shadow war is the revision, and reading the rough draft is the only way to understand what the revision is likely to achieve, what it is likely to cost, and what it is structurally incapable of resolving. The men who waited in a Roman stairwell did more than kill a translator. They wrote the opening lines of a doctrine that other states are still following, and the closing lines of that doctrine have not yet been written.

One final point belongs here, because it concerns the reader rather than the history. It is tempting, when studying an operation of this kind, to slide into one of two simple postures. The first is admiration, a fascination with the tradecraft, the patience, the audacity of a state that reached into a dozen countries and killed the men it wanted dead. The second is condemnation, a flat moral rejection of extrajudicial killing in all its forms. Both postures are too easy, and both obscure more than they reveal. The honest posture is harder to hold, because it requires keeping several true things in mind at once. It is true that the men killed in this hunt included individuals connected to the murder of innocents. It is true that the operation was conducted with remarkable skill. It is also true that it killed an innocent waiter in Norway, that it could never prove it achieved its strategic purpose, that it provoked retaliation, and that it operated entirely outside the courts of any nation. A reader who holds only the flattering truths has been seduced. A reader who holds only the damning ones has refused to engage with the genuine dilemma that produced the operation. The reason this history matters for understanding India’s shadow war is that the same discipline of holding contradictory truths together is exactly what an honest assessment of the Indian programme demands. The motorcycle assassins of Pakistan’s cities are neither simply heroes nor simply criminals. They are the latest practitioners of a doctrine that is genuinely effective at removing individuals, genuinely unable to prove it accomplishes anything strategic, and genuinely dangerous to the democratic character of the states that employ it. That is an uncomfortable conclusion, and its discomfort is the point.

Frequently Asked Questions

Q: What was Operation Wrath of God?

Operation Wrath of God was a covert Israeli campaign of assassination, also carried out under the codename Bayonet, launched in the autumn of 1972 in response to the massacre of eleven Israeli athletes and coaches at the Munich Olympics. Directed by the Israeli intelligence service and authorized by Prime Minister Golda Meir through a secret body remembered as Committee X, it sought to locate and kill the Palestinian figures believed to have planned, financed, or otherwise enabled the Munich attack and the broader campaign of terrorism around it. The programme operated mainly in Europe and the Middle East, used methods ranging from close-range shootings to telephone and car bombs to a full commando raid, and is generally understood to have continued in some form for more than two decades.

Q: How did Mossad plan and execute the campaign?

The campaign followed a consistent operational pattern. The intelligence service first assembled a target list, drawing on its own sources, Palestinian informants, and cooperation with friendly European agencies. Each name was then reviewed individually by the authorizing committee. Once approved, a target was placed under extended surveillance so that his residence, routine, and movements could be mapped in detail. Small operational cells then carried out the killing using a method chosen to suit the specific target and to avoid a predictable signature, after which the team extracted quickly to a safe house or out of the country. Plausible deniability was a structural requirement at every stage, because the strikes took place on the territory of states that had not consented to them.

Q: How many people did the programme kill?

An exact number is not known and probably never will be, because the operation was covert and its full target list has never been published. Estimates of the original list range between roughly twenty and thirty-five names. The hunt’s documented strikes include a series of well-known killings in Rome, Paris, Nicosia, Athens, and Beirut, and the Beirut commando raid known as Operation Spring of Youth killed three senior Palestinian figures along with a larger number of fighters. Killings attributed to the same campaign continued for years afterward, which makes any single total an estimate rather than a confirmed figure.

Q: What was the Lillehammer affair?

The Lillehammer affair was the operation’s defining failure. In July 1973, an Israeli team operating in the Norwegian town of Lillehammer was hunting Ali Hassan Salameh, a Palestinian operations figure Israel held responsible for the Munich massacre. The team misidentified its target and instead killed Ahmed Bouchikhi, a Moroccan-born waiter with no connection to terrorism, in front of his pregnant wife. Norwegian police arrested several members of the team, and the subsequent trial convicted five people and exposed part of the intelligence service’s European network of agents and safe houses. The scandal led Prime Minister Golda Meir to suspend the assassination programme.

Q: Who was Ali Hassan Salameh?

Ali Hassan Salameh was a senior Fatah and Black September figure whom Israel considered a principal architect of the Munich massacre. Within the intelligence service he was known as the Red Prince, and he became the hunt’s most sought-after target. He survived the Lillehammer operation, which had been aimed at him, as well as a series of further assassination attempts over the following years. He was finally killed in Beirut in January 1979 by a car bomb detonated along a route he travelled regularly, more than six years after the massacre that had set the campaign in motion.

Q: Which cities did the effort operate in?

The campaign operated across a wide geographic area. Its documented strikes took place in Rome, where the first killing occurred; Paris, the site of multiple operations including a telephone bomb and at least two further killings; Nicosia in Cyprus and Athens in Greece, both targeted with explosive devices; and Beirut, which saw both the commando raid of Operation Spring of Youth and the later car bomb that killed Salameh. The Lillehammer operation took the campaign to Norway. Killings attributed to the same broader programme in later decades extended its reach further still.

Q: Was Operation Wrath of God legal under international law?

Operation Wrath of God sits in a deeply contested area of international law, and a definitive ruling on its legality has never been issued by any authoritative body. The killings were carried out on the territory of states that had not consented and against individuals who had been convicted by no court, which conventional readings of sovereignty and due process would treat as serious violations. Defenders have argued that the operation was a form of self-defence against an ongoing terrorist threat. The Lillehammer affair, by contrast, produced an unambiguous criminal verdict in a Norwegian court. The honest summary is that the practice of extraterritorial state assassination remains legally unsettled, and this article is not a substitute for qualified legal advice on the question.

Q: How long did the campaign last?

The campaign’s most intense and concentrated phase ran from the autumn of 1972 through the summer of 1973, ending with the suspension that followed the Lillehammer disaster. It was reactivated for the killing of Salameh in 1979, and killings attributed to the same broader programme continued through the 1980s and into the 1990s, with the Paris assassination of Atef Bseiso in 1992 often cited as a late echo. In practical terms the programme is best understood not as a fixed-length operation but as a standing capability that was activated, suspended, and reactivated over a span of roughly two decades.

Q: Did the campaign deter future Palestinian attacks?

This is the central unresolved question, and the evidence does not support a confident answer. Defenders argue that the hunt imposed fear and disruption on the Palestinian network and contributed to the decline of Black September as a distinct operational entity. Critics counter that the network struck back, that killed figures were quickly replaced, and that the campaign did nothing to address the political grievances and recruitment pipelines that generated new operatives. Because the effort was covert and because the world in which it did not happen cannot be observed, the deterrence claim is essentially unfalsifiable, which means it can never be conclusively proven or disproven.

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

The parallels are extensive and not coincidental. Both campaigns were the response of a democracy to terrorist atrocities that conventional and legal remedies had failed to answer. Both faced enemies who lived openly under the protection of a state that would not extradite them. Both chose patient, methodical, deniable elimination over inaction or open war. Both built an operational signature based on extended surveillance and a strike at a predictable location, and both treated the operation as a permanent capability rather than a finite mission. India’s planners studied the Israeli model, and the intelligence relationship between the two states provided a channel for doctrine to transfer. The Israeli campaign is best understood as the historical template that India’s shadow war revises and extends.

Q: What methods did the campaign use?

The operation was deliberately varied in its methods so that no single countermeasure could defeat it. It used close-range shootings, as in the first killing in Rome and the later killing of a target in Paris. It used concealed explosive devices, including a bomb hidden inside a desk telephone, a bomb planted beneath a hotel bed, and car bombs. It also escalated, when a target’s protection demanded it, to a full amphibious commando raid in the form of Operation Spring of Youth. This variety was itself a doctrine: the unpredictability of method was a weapon as much as any individual bomb or pistol.

Q: What was Operation Spring of Youth?

Operation Spring of Youth was a military raid carried out in April 1973 as a sub-operation of the larger campaign. Several targets lived in heavily guarded apartment blocks in Beirut, beyond the reach of ordinary covert methods. In response, Israeli commandos from elite units came ashore on the Lebanese coast at night, were met by intelligence agents already in the city, and raided the target buildings. Some raiders were disguised as civilians and some as women. The raid killed three senior Palestinian figures, Muhammad Youssef al-Najjar, Kamal Adwan, and Kamal Nasser, along with a larger number of fighters. A future Israeli prime minister, Ehud Barak, took part. The raid proved that no fortified sanctuary was truly beyond reach.

Q: Why did Israel choose assassination instead of conventional military action?

Israel did initially respond conventionally, bombing Palestinian bases in Syria and Lebanon within two days of the Munich aftermath. Airstrikes, however, killed foot soldiers and destroyed buildings without reaching the planners, financiers, and organizers who made the network function. Assassination offered a more discriminating instrument: it could reach specific individuals, and it could plant fear in the entire network by demonstrating that any operative might be found and killed. The choice reflected a judgment that the programme’s true target was not Palestinian infrastructure but a specific set of people, and that those people could best be reached individually.

Q: What does the regeneration problem mean for targeted killing?

The regeneration problem is the tendency of an organization to replace an eliminated figure with a successor, sometimes one who proves harder or more radical. The campaign demonstrated this directly: when the Fatah representative in Cyprus was assassinated, a successor took the position, and the successor was then also assassinated. The lesson is that killing an individual does not kill the function he performed, because an organization is a regenerating structure rather than a fixed set of people. The regeneration problem is one of the strongest arguments against treating targeted killing as a strategy rather than a tactic, and it applies directly to any modern elimination campaign.

Q: Why does Operation Wrath of God still matter today?

It matters because it was the first complete run of a programme that is now running again. The programme established that a democracy could hunt and kill its enemies abroad, over a span of years, without admitting that it was doing so, and that the world would largely accept the practice. That precedent became an available model for other states facing similar dilemmas, and India’s shadow war is, in the most literal sense, an exercise of the permission the Israeli campaign established. Studying the original campaign, including both its operational competence and its failures, is the clearest way to understand what a modern elimination programme is likely to achieve and what it is structurally incapable of resolving.

Q: What were the costs and consequences of the campaign?

The campaign disrupted the Palestinian network in Europe, forcing survivors to divert energy toward personal security. It also exposed, through its own actions, the regeneration problem, as eliminated figures were replaced. The Lillehammer affair produced a dead innocent man, a criminal verdict, and the exposure of part of the intelligence service’s European apparatus. The hunt provoked retaliation rather than passive acceptance, and it left a permanent legal and reputational controversy. Its most lasting consequence, however, was doctrinal: it converted extraterritorial state assassination from the unthinkable into the precedented, and that precedent is now available to be followed by any state willing to follow it.

Q: Did the hunt have authorizing oversight, and how does that compare to India?

Israel’s campaign operated with at least the form of an authorizing structure. Prime Minister Golda Meir created a committee that reviewed each target individually, she took responsibility for activating the programme, and she took responsibility for suspending it after Lillehammer. In later decades the Israeli system developed a measure of judicial scrutiny over targeted killing. India’s shadow war, by contrast, is conducted in a posture of total denial, with no acknowledged authorizing process and no visible review. Whether that total deniability is an operational strength or a democratic danger is a genuine and unresolved disagreement, and it is one of the most important questions raised by comparing the two campaigns.

Q: Why is the identification problem considered the programme’s central vulnerability?

The identification problem is the risk that the wrong person is marked, tracked, and killed. It is considered central because every other element of a targeted-killing operation can function perfectly while the identification is wrong, and when that happens, the result is the death of an innocent person, as occurred at Lillehammer. The problem is structural rather than incidental. A covert programme cannot use the safeguards that a criminal court uses to test an identification before the irreversible step is taken, because the entire point of the programme is to act without that visibility. The only check left is the confidence of the officers who made the assessment, and confidence is not a reliable check. Any campaign that removes all external tests of its identifications has effectively guaranteed that some of its identifications will be wrong, and the errors will only become visible after a death.

Q: What is the regeneration problem and why does it weaken the case for targeted killing?

The regeneration problem is the tendency of an organization to replace an eliminated figure with a successor. The campaign demonstrated it directly when an assassinated Cyprus representative was replaced, and the successor then also had to be hunted and killed. The problem weakens the strategic case for targeted killing because it suggests that the effort is not steadily dismantling the enemy but is instead running on a treadmill, expending real operational risk to remove functions that the enemy immediately restores. It implies that targeted killing is a tactic, useful for removing a specific individual, rather than a strategy capable of producing the permanent defeat of an adversary.

Q: Was the campaign acknowledged by the Israeli government at the time?

No. The effort was conducted as a covert operation, and plausible deniability was a structural requirement built into every strike. The killings were designed so that no acknowledged connection to the Israeli state could be proven, even though the broader pattern made the responsible party widely suspected. Over the following decades the operation became extensively documented through journalism, memoirs, and historical research, and Israel’s posture shifted toward a kind of semi-acknowledgment of its targeted-killing practice in general. At the time of the operations themselves, however, the official position was silence, and that silence was itself part of the operational design.

Q: How is this history relevant to readers following India and Pakistan today?

It is relevant because India’s shadow war is, in its structure and method, a direct descendant of the Israeli campaign. A reader who wants to understand why terrorists are being killed by unidentified attackers in Pakistani cities, why the Indian government never acknowledges the killings, why the campaign shows no sign of ending, and why its actual effectiveness is so difficult to assess will find that every one of these questions was already posed, and partially answered, by Operation Wrath of God decades earlier. The Israeli campaign is the rough draft of the programme India is now running. Reading the rough draft, including its successes, its catastrophic failure at Lillehammer, and the strategic questions it could never resolve, is the clearest way to understand what India’s campaign is likely to achieve and what it is structurally unable to settle.

Q: What role did false documentation and cover identities play in the operations?

Cover identities were not a peripheral convenience but a load-bearing element of every strike. The operatives who moved through Rome, Paris, and Nicosia travelled on documents that presented them as tourists, businessmen, and journalists from a range of European countries, and the credibility of those documents determined whether a team could rent a car, occupy a hotel, observe a target for days, and depart without ever registering as a threat. The false journalist who arranged the fatal telephone interview with Mahmoud Hamshari is the clearest illustration of how a cover identity could be turned from a passive shield into an active instrument of the kill itself. The same dependence on credible cover later became a vulnerability. When the Lillehammer team was arrested, the documents and rental records that should have protected the operatives instead became the thread that investigators pulled to unravel a wider European network. India’s planners inherited the same double-edged tool, because a deniable elimination depends entirely on operatives who can arrive, watch, strike, and vanish without ever being recorded as what they are.

Q: Did the targeted individuals know they were being hunted, and did it change their behaviour?

Yes, and the change in their behaviour is itself one of the most revealing findings of the whole history. As the strikes accumulated, the surviving names on the list understood that they had become targets, and they responded by altering their routines, increasing their protection, varying their movements, and treating ordinary activities such as answering a telephone or returning home at a fixed hour as potential points of danger. Ali Hassan Salameh, the most wanted figure of all, converted his own survival into a disciplined exercise that imposed roughly six years of delay on one of the world’s most capable intelligence services. This adaptation cuts in two directions for any assessment of targeted killing. On one hand, it confirms the deterrence argument, because an operative who is distracted, cautious, and consumed by his own security is an operative who is less effective at planning attacks. On the other hand, it raises the cost and difficulty of every subsequent strike, because a hunted man who has learned he is hunted is a far harder target than one who still believes himself safe. India’s eliminations in Pakistan operate against the same dynamic, where each successful strike teaches the remaining network to be more careful and therefore harder to reach.