On a quiet residential street in Coyoacan, a leafy borough of Mexico City, an exiled revolutionary sat at his desk reading a manuscript handed to him by a young man he believed was a sympathizer. The visitor stood behind him, drew a shortened mountaineering axe from beneath his raincoat, and brought it down on the back of the reader’s skull. The victim screamed, fought back, and lived through the night before dying the following day. The man at the desk was Leon Trotsky, co-architect of the Bolshevik Revolution and Joseph Stalin’s most hated rival. The killing of August 1940 was not a crime of passion or a private vendetta. It was a state operation, planned over years, funded from Moscow, and executed by an apparatus whose descendants would still be staging similar killings on foreign soil eight decades later.
That apparatus has changed its name many times. It was the Cheka, then the OGPU, then the NKVD, then the MGB, then the KGB, and after 1991 it fractured into the Federal Security Service and the Foreign Intelligence Service while the military side, the GRU, continued as it had since the Russian Civil War. The acronyms are a confusion that Moscow has, at times, found useful. What did not change across those generations of letterhead was a particular conviction at the heart of the Soviet and then Russian security state: that the physical elimination of enemies abroad is a legitimate, sometimes necessary, instrument of policy, and that the act of elimination is most valuable when the world can guess who ordered it. This is the thread that runs from the ice axe in Coyoacan through a ricin pellet fired from an umbrella on a London bridge, through a teapot laced with a radioactive isotope, to a nerve agent smeared on a suburban door handle in an English cathedral city. The methods evolved from the crude to the exotic to the industrial. The underlying logic stayed remarkably constant.
This is the story of that thread. It is a history of operations, of the men who planned and carried them out, and of the institutional culture that produced them. It is also an argument about why a state that has every technical means to kill quietly so often chooses to kill loudly, and what that choice reveals about how Moscow understands fear, deterrence, and the management of its own diaspora of dissidents and defectors. For readers following the wider question of how states reach beyond their borders to settle accounts, the Soviet record offers the longest continuous case study available. It is older than the Israeli campaign that grew out of the Munich massacre, older than the American drone program, and older than India’s own contested record of eliminations on foreign soil. Understanding it is a prerequisite for understanding all the others.
Background and Triggers: The Bolshevik Birth of State Killing
The willingness to kill abroad was not a late corruption of the Soviet intelligence services. It was present at their creation. The Cheka, the All-Russian Extraordinary Commission, was founded in December 1917, weeks after the Bolsheviks seized power. Its first chairman, Felix Dzerzhinsky, gave it a mandate to defend the revolution by any means, and the men who built it understood that the revolution had enemies who would not stay conveniently inside Russian borders. Aristocrats, officers of the defeated White armies, monarchists, socialist rivals, and nationalist leaders fled by the hundreds of thousands into Europe after the Civil War. They formed governments in exile, published newspapers, plotted returns, and lobbied foreign powers. Moscow regarded this emigration not as a spent force but as a standing threat, a reservoir of counter-revolution that might at any moment be funded and armed by hostile capitals.
Defending the revolution against that reservoir required, in the early years, more cunning than raw firepower. The most famous operation against the exiles, the Trust, was an elaborate fiction in which the Soviet security service ran a fake anti-Bolshevik underground, luring exiled enemies back across the border into captivity. Kidnapping followed manipulation as the favored instrument. Two successive heads of the main White veterans’ organization in Paris were abducted, General Alexander Kutepov in 1930 and General Yevgeny Miller in 1937, in operations that showed how far the service would reach. Kutepov was seized on a Paris street and is believed to have died during the journey to Soviet territory. Miller was lured to a meeting, drugged, packed into a crate, and shipped home to be interrogated and shot. These were not assassinations in the narrow sense, but they established a principle that mattered for everything that followed. The territory of another sovereign state was not, in Moscow’s operational thinking, a barrier. It was simply terrain.
A shift from kidnapping to outright killing tracked the rise of Stalin and the logic of his terror. Inside the Soviet Union, the purges of the 1930s consumed Old Bolsheviks, military officers, and millions of ordinary citizens. Abroad, the same paranoid logic demanded that anyone who had broken with Stalin and knew the secrets of the apparatus be silenced. The defection of intelligence officers was treated as the gravest possible offense, because a defector carried in his head the identities of agents, the methods of operations, and the embarrassing truth about how Moscow conducted its affairs. The case of Ignace Reiss is the clearest early example. Reiss was a Soviet intelligence officer who broke with Stalin in 1937, denounced the purges, and declared his loyalty to the exiled Trotsky. Within weeks a mobile group tracked him to Switzerland, ambushed his car near Lausanne, and machine-gunned him on a country road. The message to every other officer who might be tempted to defect was delivered in lead.
That same year produced an operation that revealed the technical ambition already present in the service. Yevhen Konovalets, the leader of the Organization of Ukrainian Nationalists, was a serious adversary, a man capable of mobilizing Ukrainian sentiment against Soviet rule. The officer assigned to remove him, Pavel Sudoplatov, spent months cultivating Konovalets as a trusted contact, posing as a sympathizer from within the Soviet Ukraine. In May 1938, in a Rotterdam restaurant, Sudoplatov handed the nationalist leader a gift, a box of chocolates that was in fact a bomb. He left, the device detonated, and Konovalets was killed. Sudoplatov walked away and went on to become the single most important figure in the Soviet history of targeted killing, the man who would later organize the hunt for Trotsky himself. The Konovalets killing is worth holding in mind because it demonstrates how early the service mastered the disguised weapon, the device that looks like an ordinary object and turns lethal in the victim’s hands. The umbrella, the perfume bottle, and the contaminated teacup were all decades in the future, but the conceptual lineage was already set.
Worth emphasizing, too, is the bureaucratic ordinariness with which this work was administered. The men who organized killings abroad were not freelance fanatics operating outside the system. They were salaried officers who filed reports, requested resources, attended planning meetings, and were promoted or demoted on the strength of their results. A successful elimination, inside the institution, was simply a closed file, a problem solved and recorded. That mentality, the treatment of lethal action as an administrative function of statecraft, is the cultural inheritance that every successor agency received, and it goes a long way toward explaining why the practice proved so durable across changes of regime, ideology, and acronym. An apparatus does not easily abandon a tool it has learned to use without guilt. It is worth noting, as well, how the early operations distributed risk among different kinds of people. The mobile groups that ambushed Ignace Reiss, the recruited assets who staged kidnappings on Paris streets, the deep-cover officer who befriended Konovalets for months before delivering the bomb: each model placed a different sort of human being at the point of danger, and each taught the planners something about which arrangements held under pressure and which collapsed. By the close of the 1930s the service had, in effect, run a decade-long experiment in how to kill at a distance, and it had accumulated the institutional knowledge to do so with growing confidence.
By the end of the 1930s, then, the Soviet security apparatus possessed a settled doctrine and a proven capability. The doctrine held that enemies of the state forfeited the protection of distance and foreign sovereignty. The capability included surveillance networks across the European emigration, mobile groups trained for ambush, and a willingness to invest in disguised and exotic weapons. What the apparatus needed was a target important enough to justify the most ambitious operation it had ever attempted. Stalin supplied one. The target was the man who, more than any other living person, embodied the idea that there was an alternative Bolshevism, a revolution that Stalin had betrayed. That man was Trotsky, and the campaign to kill him would become the template against which every later operation can be measured.
The Trotsky Operation: An Ice Axe in Coyoacan
Leon Trotsky was not merely a political opponent. To Stalin he was an existential reproach. Trotsky had led the Red Army to victory in the Civil War, had been spoken of as Lenin’s natural successor, and had spent his exile producing a torrent of writing that portrayed Stalin as a gravedigger of the revolution. Expelled from the Soviet Union in 1929, hounded from Turkey to France to Norway, Trotsky finally found refuge in Mexico in 1937, settling first in the home of the painter Diego Rivera and later in a fortified house in Coyoacan. From there he organized the Fourth International, the rival to Stalin’s Communist movement, and continued to write the biography of his nemesis. For Moscow, the existence of an organized, articulate, internationally visible alternative leadership was intolerable. The order to eliminate him was given, and the operation received the codename Utka, the Russian word for duck.
Sudoplatov was placed in overall charge. The operation he designed was layered, with two independent networks that did not know of each other, so that the failure or compromise of one would not doom the other. The first network was built around the Mexican muralist David Alfaro Siqueiros, a committed Stalinist and a man of action. In the pre-dawn hours of May 24, 1940, a group of some twenty men in police and army uniforms, led by Siqueiros, overpowered the guards at the Coyoacan house, burst into the compound, and raked the bedrooms with submachine-gun fire. They fired hundreds of rounds. And they failed. Trotsky and his wife Natalia survived by rolling into a corner of their bedroom and lying flat against the wall as the bullets passed above them. Their grandson was grazed but lived. The attackers, believing the job done, withdrew. The crudeness of the raid, its reliance on volume of fire and on the assumption that a hail of bullets into a dark room would be sufficient, marks one end of the methodological spectrum this history will trace.
The second network was the one that succeeded, and it succeeded because it relied not on force but on intimacy. Its instrument was a young Spaniard named Ramon Mercader, the son of Caridad Mercader, a fervent Communist and herself an agent of the service. Mercader had been cultivated for years. Using a false identity, posing variously as a Belgian businessman and traveling on a doctored Canadian passport under the name Frank Jacson, he insinuated himself into the circle around Trotsky by courting an American woman who was part of the household’s support network. Over months he became a familiar, trusted figure, a man who could come and go, who brought small gifts, who asked the great revolutionary for help with his writing. He was, in the language of the trade, a penetration agent who had achieved the deepest possible access. He could stand behind Trotsky’s chair without anyone in the house feeling alarm.
Patience of that kind is the operation’s real lesson, and it is worth dwelling on. The Coyoacan house was a fortress by 1940, with high walls, watchtowers, steel doors, an alarm system, and a permanent guard detail of devoted followers. None of it mattered. The fortifications were built to stop an assault, and the second network never mounted an assault. It simply walked a trusted friend in through the front gate. The contrast between the two approaches, the violent raid that the defenses were designed to defeat and the quiet infiltration that the defenses were powerless against, is the single most instructive feature of the whole campaign. It established, for every planner who came after, that the surest route to a guarded target runs not through the wall but through the target’s own circle of trust. A bodyguard can stop a stranger with a gun. A bodyguard cannot stop a friend with a manuscript. That insight, learned at such cost in Mexico, would echo through the tradecraft of the decades that followed, shaping how the service thought about cultivation, cover, and the slow construction of access.
On August 20, 1940, Mercader arrived at the house carrying a manuscript he wanted Trotsky to review, a draft article. He also carried, concealed, three weapons: a pistol, a dagger, and a shortened ice axe, the kind used by mountaineers, with most of the handle sawn off so it could be hidden under a coat. While Trotsky bent over the manuscript at his desk, Mercader struck him once on the head with the axe. The blow was savage but not instantly fatal. Trotsky cried out, struggled to his feet, and fought his assailant, biting his hand. Guards rushed in and seized Mercader, who would have been killed on the spot had Trotsky not, even in his agony, ordered them to keep the man alive so that he could be interrogated and the truth of who sent him established. Trotsky was taken to a hospital, underwent surgery, and died the following day, August 21, 1940. He was sixty years old.
Mercader served twenty years in a Mexican prison. Throughout his trial and imprisonment he maintained a cover story, claiming to be a disillusioned follower acting on personal motives, and he never publicly admitted his connection to Soviet intelligence. His mother was spirited out of Mexico. When Mercader was released in 1960 he traveled to the Soviet Union, where he was secretly awarded the title Hero of the Soviet Union, the state’s highest honor, presented for what the citation could only describe in veiled terms. Sudoplatov, the organizer, was decorated as well. The Coyoacan operation thus closed the loop that the Konovalets killing had opened. The state had identified its most important enemy, had invested years and two parallel networks in his destruction, had accepted the failure of the crude approach and pressed on with the patient one, and had then quietly rewarded the men responsible. The killing of Trotsky was, in every sense, a model. It demonstrated that no level of fame, no distance, no foreign refuge, and no fortified house could place a target beyond Moscow’s reach. Every subsequent operation in this history is, in one way or another, a variation on that demonstration.
It is worth pausing on a detail that the official Soviet response made unmistakable. For decades Moscow denied any role in the killing, and Mercader’s cover story, the tale of a disillusioned follower acting alone, was the public face of the operation. Yet the secret award of the Soviet Union’s highest honor, presented once the assassin reached safe territory, tells the real story. The state denied the deed in public and rewarded it in private, and the two acts were not in tension. They were complementary halves of a single doctrine. Public denial preserved the formal fiction that the Soviet Union did not murder its enemies abroad; private reward ensured that the men who carried out such work understood the state stood behind them. That arrangement, denial outward and recognition inward, would be repeated, with variations, in case after case across the decades that followed, and it survived the Soviet collapse intact. The pattern visible in the Litvinenko aftermath, official rejection of the findings paired with the protection and even the political elevation of a suspect, is the same arrangement that first appeared in the handling of Mercader. The continuity is not a coincidence. It is the institutional memory of a state that learned, in Mexico, how to own a killing and disown it at the same time.
SMERSH, the Special Bureau, and the Defector Hunts
The Second World War transformed the Soviet apparatus of killing from an instrument aimed mainly at exiles into a vast machine. The wartime counterintelligence organization SMERSH, an acronym formed from the Russian phrase meaning death to spies, became legendary for its ruthlessness in hunting traitors, collaborators, and enemy agents within and behind the advancing Red Army. SMERSH belongs to this history partly because of its reputation and partly because it folded into the postwar services a generation of officers for whom the summary elimination of state enemies was simply routine. After the war, the function of organizing assassinations abroad was concentrated in a specialized unit, often referred to as the Special Bureau or, in later years, by department numbers that changed with each reorganization. Sudoplatov ran it for much of this period. The unit maintained its own laboratory for the development of poisons and disguised weapons, an entity known informally as the Kamera, the chamber, where toxins were tested and lethal devices engineered.
Conditions of the Cold War supplied this machine with a steady stream of targets, and those targets fell into two broad categories. The first was the nationalist leadership of the captive nations, above all Ukraine, whose emigre organizations continued to agitate against Soviet rule from bases in West Germany and elsewhere. The second was the growing population of defectors, intelligence officers and officials who had crossed to the West and whose continued existence was both a security risk and a standing humiliation. Each category demanded a slightly different operational approach. The nationalist leaders were public figures who could be located through open political activity, while the defectors often lived under protection and changed their names, requiring patient tracking. What united the two categories was Moscow’s conviction that neither distance nor a new identity should be permitted to confer immunity. The euphemism the service used for the lethal work was mokroye delo, wet affairs, a phrase that captured the bureaucratic chill with which the killing was administered. Wet affairs were authorized at the highest levels, planned by professionals, and entered into the institutional memory as solved problems.
A clearest illustration of how the postwar machine operated is the career of a single assassin, Bohdan Stashynsky. A young Ukrainian recruited by the Soviet service, Stashynsky was trained for a very particular kind of work. His weapon was not an axe or a gun in the ordinary sense. It was a device that sprayed a jet of poison gas, a crushed cyanide capsule whose vapor, inhaled at close range, would induce what looked like a heart attack and leave little forensic trace. In October 1957, in Munich, Stashynsky used this device to kill Lev Rebet, a Ukrainian nationalist writer and political figure. Rebet collapsed on a staircase and was assumed for years to have died of natural causes. The operation was, by the service’s standards, a complete success, because the victim was dead and the cause appeared innocent.
Two years later Stashynsky was sent against a far bigger target. Stepan Bandera was the most prominent and most divisive figure in the Ukrainian nationalist emigration, a man Moscow regarded as a deadly enemy and as a symbol around which anti-Soviet Ukrainian sentiment dangerously coalesced. In October 1959, in the entrance hall of his Munich apartment building, Stashynsky killed Bandera with the same cyanide spray device. Once again the death was at first attributed to natural or accidental causes, and once again Stashynsky was decorated by the Soviet state for his work. Here the postwar pattern reaches its purest expression: a trained specialist, a custom poison weapon designed to disguise murder as misfortune, a high-value political target, and a quiet reward. It is the doctrine of the deniable killing, the operation designed so that no one can prove a state did anything at all.
And then the pattern broke, in a way that would shape Soviet thinking for a generation. Stashynsky, troubled by what he had done and influenced by his East German wife, defected to the West in 1961. He confessed in detail to both killings, described the spray device, named his handlers, and was tried publicly in West Germany in 1962. The court, in a judgment that reverberated through the intelligence world, treated Stashynsky as in important respects a tool of the state that had directed him and handed down a relatively limited sentence, while placing the moral and legal weight of the murders on the Soviet leadership that had ordered them. The Stashynsky affair was a disaster for Moscow. It converted two deniable killings into a public, documented indictment of the Soviet state as a murderer of exiles. The lesson the service drew was cautious. For a period the leadership grew wary of close-quarter assassinations of emigres precisely because the Stashynsky case had shown how a single defecting assassin could expose the entire program. The episode is a crucial hinge in this history. It is the moment the Soviet apparatus learned, painfully, that the deniable killing is only deniable until the man who did it decides to talk.
Consider what that episode actually proved, because it sits at the heart of every operational dilemma in this history. A deniable killing depends on a chain of human beings, and the chain is only as strong as its least reliable link. Mercader had held his cover for twenty years inside a Mexican prison; Stashynsky broke his within two years and brought a courtroom’s worth of detail with him. The same service, the same doctrine, and the same kind of training produced both men, and the difference between them came down to character, conscience, and circumstance. After 1962 the leadership understood, in a way it had not before, that the perfect untraceable operation carries inside it a permanent vulnerability, namely the operative himself. That understanding did not end the program. It did push the service, over the following decades, toward arrangements that reduced its direct exposure.
One such arrangement deserves particular attention, because it carried the doctrine forward into the era of the umbrella. Rather than always sending its own officers, the Soviet service increasingly preferred to lend its lethal expertise, its poisons and its engineering, to the allied intelligence organs of the Warsaw Pact. The junior partner supplied the operatives and absorbed the political risk; the senior partner kept the technological monopoly and a layer of distance. This was not a retreat from the practice of killing abroad. It was a restructuring of how the killing was done, designed to insulate Moscow from the kind of exposure that Stashynsky had inflicted. The next major case in this history, the killing of a Bulgarian writer on a London bridge, is the clearest possible demonstration of how that restructured model worked in practice.
The Markov Umbrella: Ricin on Waterloo Bridge
If the Stashynsky affair taught Moscow caution about doing the work directly, it also pointed toward a solution that the Cold War’s network of allied intelligence services made possible. The Soviet security establishment sat at the center of a bloc. The Warsaw Pact states each maintained their own security organs, and those organs were trained, equipped, and in important respects directed by the apparatus in Moscow. When a member of the bloc needed a difficult job done, the senior partner could provide the technology and the expertise while the junior partner provided the operatives and absorbed the political exposure. The killing of Georgi Markov is the textbook case of this arrangement, and it produced the single most famous murder weapon in the history of intelligence: the poison umbrella.
Georgi Markov was a Bulgarian writer of real stature, a novelist and playwright who had been close to the cultural establishment of Communist Bulgaria before breaking with the regime and defecting to the West in 1969. He settled in London, where he worked as a broadcaster for the BBC World Service and for Radio Free Europe, and where his commentaries dissected the Bulgarian leadership with an insider’s knowledge and a satirist’s contempt. To the Bulgarian regime, and in particular to its long-ruling leader Todor Zhivkov, Markov was an intolerable irritant, a defector whose voice carried into Bulgarian homes and whose mockery undermined the dignity of the state. The decision to silence him was taken in Sofia, but the means were supplied by Moscow. The Soviet service provided the toxin, ricin, a poison derived from the castor bean that is lethal in minute quantities and for which there is no antidote, and it provided the engineering for a delivery system disguised as an ordinary object.
On the morning of September 7, 1978, Markov walked across Waterloo Bridge toward a bus stop on his way to work. In the crowd he felt a sharp jab in the back of his right thigh. He turned and saw a man stooping to pick up an umbrella; the man mumbled an apology in a foreign accent, crossed the road, and hailed a taxi. Markov thought little of it at first. By that evening he was running a fever. He was admitted to hospital, told the doctors about the strange incident with the umbrella, and his condition deteriorated rapidly as his body went into systemic failure. Georgi Markov died on September 11, 1978, four days after the jab. He was forty-nine years old.
It was the post-mortem that turned a suspicious death into one of the defining cases of Cold War espionage. Pathologists recovered from the wound in his thigh a tiny metal sphere, smaller than the head of a pin, drilled with two minute cavities. The pellet had been engineered to carry a microscopic dose of ricin, sealed with a coating designed to dissolve at body temperature so that the poison would be released only after the sphere was inside the victim. The umbrella, investigators concluded, had been modified into a pneumatic device capable of firing the pellet a short distance into a target brushed against in a crowd. The forensic reconstruction was supported by a near-identical case. Some days earlier, a Bulgarian defector named Vladimir Kostov had felt a similar jab in a Paris metro station. Kostov fell ill but survived, and when doctors, alerted by the Markov case, examined him, they recovered an identical pellet from his back. The coating on Kostov’s pellet had not fully dissolved, which is almost certainly why he lived. Two pellets, two cities, one weapon, one method. The Bulgarian umbrella was no longer a theory.
The division of labor in the Markov case is the analytically important point. The political decision belonged to Sofia. The operatives were Bulgarian, and the man on the bridge has never been conclusively identified and convicted. But the toxin, the pellet technology, and the technical training came from the Soviet service, whose laboratory had the expertise that the Bulgarian organs did not. This is the bloc model of assassination at work, and it served Moscow’s purposes precisely. The killing advanced the interests of a Soviet client, demonstrated to every defector that the reach of the bloc extended into the heart of a Western capital, and yet kept the Soviet service one step removed from the act itself. The umbrella belongs to this history not because the KGB pulled the trigger but because the KGB built the gun. It is the clearest single illustration of how the Soviet apparatus could project lethal force while preserving a layer of deniability, by lending its technology to the services it had itself created.
Reflect for a moment on why that arrangement was so attractive to the senior partner. Moscow gained the deterrent value of the killing without bearing the full diplomatic cost of having carried it out. Should the operation be exposed, the trail led first to Sofia, and the Soviet contribution, the laboratory work and the training, sat behind a screen of plausible distance. The bloc model was, in effect, an institutional answer to the vulnerability the Stashynsky affair had revealed. The Markov case is also worth studying for the sheer ingenuity of the weapon, which represented the high-water mark of the disguised-device tradition. A working umbrella that doubled as a pneumatic gun, a pellet drilled to tolerances measured in fractions of a millimeter, a sugar coating engineered to dissolve at the precise temperature of the human body: every element was designed so that the victim would register nothing more than a clumsy stranger and a momentary sting. Had the coating on Markov’s pellet behaved as the coating on Kostov’s pellet did, the writer might have survived, and had he died of the systemic failure without the pellet ever being found, his death might have been recorded as an obscure illness. Concealment, in 1978, was still the governing ambition, and the umbrella was concealment raised to the level of craftsmanship.
There is a further dimension to the Markov killing that the focus on the weapon can obscure, and it concerns the nature of the target. Markov was not an intelligence officer carrying operational secrets. He was a writer and a broadcaster, a man whose only weapon was his voice and his pen. What made him intolerable to the regime in Sofia was not what he knew but what he said, the steady, mocking commentary that reached Bulgarian listeners and stripped the leadership of its dignity. His killing therefore widened the category of who could be a target. The defector hunts had been aimed at people who posed a security threat in the conventional sense, men whose knowledge could damage operations. Markov posed a threat of a different kind, the threat of speech, and his elimination signaled that the bloc would reach across borders to silence not only those who could betray its secrets but those who could merely embarrass it. That widening matters for the cases that came later. Several of the modern Russian targets, including some of the most prominent critics living in exile, were men whose offense, like Markov’s, was primarily one of speech. The umbrella killing established, in 1978, that an articulate voice was reason enough.
The Litvinenko Case: Polonium in a London Teapot
The collapse of the Soviet Union in 1991 dismantled the KGB as an institution but did not dismantle the doctrine. The agency was divided. Its foreign intelligence directorate became the SVR, the Foreign Intelligence Service. Its domestic security functions became the FSB, the Federal Security Service. The military intelligence service, the GRU, carried on unbroken. Through the 1990s, amid the chaos and impoverishment of post-Soviet Russia, the lethal capability seemed to recede. It returned, in the new century, with a new confidence, and the case that announced its return was the most technologically extraordinary assassination of the modern era.
Alexander Litvinenko was a product of the system that killed him. He had been an officer of the FSB, working in units that dealt with organized crime and counterterrorism. In the late 1990s he went public with explosive allegations against his own service, claiming among other things that he had been ordered to assassinate the businessman Boris Berezovsky and alleging FSB involvement in crimes against Russian citizens. Prosecuted and harassed in Russia, Litvinenko fled in 2000 and was granted asylum in the United Kingdom, where he became a British citizen, worked as a consultant and writer, associated with prominent Kremlin critics, and continued to make grave accusations about the conduct of the Russian state and its leadership. He was, in the long lineage this history traces, a classic target: a defector from the security service itself, in possession of damaging knowledge, who would not stay silent.
On November 1, 2006, Litvinenko met two Russian men, Andrei Lugovoi and Dmitry Kovtun, at the Pine Bar of the Millennium Hotel in the Mayfair district of London. Lugovoi was a former officer of the KGB and the FSB; Kovtun had a background connected to the GRU. The three drank tea. Within hours Litvinenko was violently ill, and over the following days his condition baffled the physicians treating him. His hair fell out. His organs began to fail. It was only as he was dying that tests identified the cause: he had ingested polonium-210, a rare and ferociously radioactive isotope. Polonium-210 emits alpha radiation, which is harmless outside the body, easily stopped by skin or a sheet of paper, but catastrophic once swallowed. A quantity smaller than a grain of salt is enough to kill. Alexander Litvinenko died on November 23, 2006, the first known victim of deliberate polonium poisoning. In a statement dictated from his deathbed he accused the Russian president by name of responsibility for his murder.
What followed was a forensic investigation without precedent. Because polonium-210 leaves a radioactive signature, scientists were able to map the movements of the men who had carried it. Traces were found at more than forty locations across London and beyond: in hotel rooms, in restaurants, in a sports stadium, and aboard aircraft that had flown between London and Moscow. The contamination trail revealed that the poisoners had made earlier, failed attempts before the successful one at the Pine Bar. The highest reading of all came from the teapot. The British public inquiry, chaired by a senior retired judge, examined this evidence over months and reported its conclusions in January 2016. The inquiry found that Litvinenko had been deliberately poisoned by Lugovoi and Kovtun, that the two men had been acting on behalf of others, that there was a strong probability they were directed by the FSB, and, in its most consequential single sentence, that the operation to kill Litvinenko had probably been approved at the highest levels of the Russian state. Russia rejected the findings and has consistently refused to extradite the two suspects, one of whom went on to a career in Russian politics. In 2021 the European Court of Human Rights, examining the same evidence, found Russia responsible for the killing.
What raised the Litvinenko case above the level of an ordinary intelligence killing was the sheer modernity of the forensic confrontation it produced. Earlier operations in this history had been investigated, where they were investigated at all, by national police forces working largely alone, often years after the fact, and frequently against a wall of official silence. The Litvinenko investigation was different in kind. It mobilized nuclear scientists, radiation specialists, and counterterrorism detectives; it produced a contamination map of a Western capital; and it culminated in a formal public inquiry with the legal power to examine secret intelligence material and to compel testimony. The result was a documented, judicially endorsed finding that pointed at the leadership of a permanent member of the United Nations Security Council. Nothing comparable had happened to any earlier operation in the Soviet record. The defector hunts of the 1950s had produced, at most, a single criminal trial of a single defecting assassin. The Litvinenko case produced an institutional verdict. That escalation, from a death that could be denied into a death that a foreign state had to spend years formally contesting, is itself a measure of how the modern operation differs from its Cold War ancestors. The weapon had changed, and so had the world’s capacity to read what the weapon left behind.
The Litvinenko case marks a turning point in this history because of what the choice of weapon reveals. Polonium-210 is not an off-the-shelf poison. It is produced in nuclear reactors. The quantity used to kill Litvinenko, scientists concluded, pointed to a state facility as its source. A service that wished simply to make a defector dead and to leave no trace had, by 2006, a wide menu of options. Polonium was not chosen for stealth; it is, after the fact, one of the most traceable substances imaginable, and it led investigators straight back along the contamination trail. It was chosen, the logic of the case suggests, for other reasons. It killed slowly and agonizingly, over three weeks of public dying that the world’s media documented in detail. It required resources only a state could command. And it delivered, to anyone paying attention, a message about capability and impunity that an ordinary bullet could never carry. The polonium operation looks, in retrospect, less like a quiet elimination than like a demonstration. That distinction, between the killing meant to be hidden and the killing meant to be understood, sits at the center of the analytical debate this article will reach.
Notice, too, the secondary effect that the choice of polonium produced, an effect that no ordinary poison could have generated. The radioactive trail did not merely incriminate the two suspects; it turned a wide swath of central London into a contamination map and forced a substantial public health response, with thousands of people contacted and tested for possible exposure. The killing of one man became, briefly, a civic emergency in a foreign capital. Whether or not that consequence was deliberately intended, it amplified the message far beyond the death itself. A polonium poisoning is not a discreet act that the host country can quietly absorb and forget. It is an event that compels the host government to acknowledge, publicly and at length, that a foreign state has reached into its territory and used a radiological substance against a person living there. The contrast with the Markov killing is instructive. Markov’s death, for all its drama, could in principle have passed as an illness; the pellet had to be found before anyone knew a crime had occurred. Litvinenko’s death announced itself. From the moment the isotope was identified, there was no version of the story in which the Russian state was not the obvious suspect, and the years of investigation and the eventual judicial inquiry only confirmed what the choice of weapon had already broadcast.
The Salisbury Attack: Novichok in a Cathedral City
Twelve years after the polonium case, the doctrine produced its most reckless expression to date, and it did so against a target who had, in a sense, already been dealt with once before. Sergei Skripal was a former colonel of the GRU, Russian military intelligence, who had been recruited by British intelligence and had passed secrets to London over a period of years in the 1990s and early 2000s. He was caught, tried in Russia, and imprisoned. In 2010 he was released as part of a large spy swap, an exchange of intelligence personnel of the kind that has long been an accepted, almost civilized, mechanism of the espionage world. Skripal settled quietly in the cathedral city of Salisbury in southern England. The spy swap carried with it an implicit understanding, honored for decades by all sides, that an officer exchanged in such a deal had been accounted for and would be left alone. The attack on Skripal tore up that understanding.
That breach is worth weighing on its own terms, because it represents a kind of damage distinct from the death and injury the operation caused. The exchange of captured intelligence personnel is one of the few genuinely stabilizing conventions in the otherwise lawless world of espionage. It works only because every service trusts that an officer handed over in a swap is closed business, that the account has been settled and will not be reopened. An operative considering whether to cooperate with a foreign service, knowing that capture might one day be followed by exchange, relies on that convention for a measure of safety. By poisoning a man who had been swapped eight years earlier, the operation did more than attempt one killing. It signaled that the convention itself could be discarded whenever Moscow chose, and that no past settlement conferred lasting immunity. The message, if it was a message, reached every intelligence officer in the world who might ever weigh defection or cooperation against the hope of an eventual exchange. Whether the planners intended that broader effect or simply did not value the convention enough to preserve it, the result was the same: a norm that had quietly reduced the lethality of the spy world for decades was shown to be conditional on Moscow’s continued consent.
On March 4, 2018, Sergei Skripal and his adult daughter Yulia, who was visiting from Russia, were found slumped and unconscious on a public bench in the center of Salisbury. They had been poisoned with a nerve agent of a type developed by the Soviet Union under the program name Novichok, a class of chemical weapon among the most toxic ever synthesized. The agent had been applied, investigators determined, to the front door handle of Skripal’s home. Both Skripals, against the expectations of the doctors who first treated them, survived, after weeks of critical illness. A police officer who attended the scene was also seriously poisoned and recovered. The Organisation for the Prohibition of Chemical Weapons, the international body, independently confirmed the identification of the nerve agent. The British government concluded that the attack had been carried out by officers of the GRU.
The Salisbury operation then produced a consequence that exposes, more starkly than any other single fact in this history, the danger of the chosen method. The two officers who had traveled to Salisbury are believed to have brought the nerve agent in a counterfeit perfume bottle, and to have discarded it after the attack. Several months later, in a town a few miles from Salisbury, a local man found what he took to be a discarded bottle of perfume and gave it to his partner, a woman named Dawn Sturgess. She sprayed it on her wrists. Dawn Sturgess fell gravely ill and died on July 8, 2018. She had no connection to espionage, to Russia, or to Sergei Skripal. She was an English woman killed by a weapon of war left lying in a public place by the operatives of a foreign intelligence service. Her death is the irreducible fact against which any claim about the precision or professionalism of the operation must be measured.
The aftermath compounded the recklessness with farce. British investigators, working with open-source researchers, identified the two GRU officers, who had traveled under the false names Alexander Petrov and Ruslan Boshirov. The men appeared on Russian television to deny everything, offering an account in which they were merely tourists who had traveled to Salisbury, twice, in winter, to admire the cathedral and its famous spire. The interview was met with near-universal disbelief and became an object of ridicule. Investigative journalists subsequently published what they presented as the officers’ real identities and military intelligence backgrounds. The Salisbury case thus combined three things that the older operations in this history had largely kept separate: a chemical weapon of mass-casualty potential deployed in a residential English town, the death of an entirely innocent bystander, and a cover story so transparent that it insulted the intelligence of everyone who heard it.
It is worth being precise about what made the choice of a nerve agent so different from anything that had come before. The cyanide spray of the 1950s and the ricin pellet of 1978 were single-victim instruments. They were dangerous to the operative and to the intended target, and to almost no one else. Novichok is not a single-victim instrument. It is a persistent chemical agent designed for battlefield use, and applying it to a door handle in a residential street meant placing a small reservoir of one of the deadliest substances ever synthesized in a public space, where weather, passersby, and pure chance would govern what happened next. The death of Dawn Sturgess was not a freak accident that defied the logic of the operation. It was a possibility built into the logic of the operation from the moment the method was selected. A weapon of that nature does not distinguish between a former intelligence officer and a woman who finds a discarded bottle months later. The older tradition, whatever else can be said against it, had at least confined its lethality to the person it was aimed at. The Salisbury method abandoned even that minimal discipline. The discussion of this attack connects directly to the broader treatment of Russia’s use of poison on British soil, and the question it forces is whether such an operation represents a failure of tradecraft or the deliberate adoption of a method whose very crudeness is the message.
The Methodology Evolution: From Ice Axe to Nerve Agent
Lay the principal operations of this history side by side and a striking trajectory emerges in the choice of weapon. It is worth tracing deliberately, because the evolution of the method is itself a form of evidence about the evolution of the institution’s priorities.
Killing Trotsky in 1940 sits at the crude end of the spectrum. The first attempt was a mass of gunmen firing hundreds of rounds into a darkened house, an approach that depended on overwhelming volume and produced overwhelming failure. The second, successful attempt used an ice axe, a tool, a brutally physical instrument that required the assassin to be within arm’s reach of his victim, to strike with his own hand, and to be immediately seized. There was nothing subtle about the ice axe. Its only sophistication lay in the years of patient penetration that placed Mercader behind the chair. The weapon itself was as old as human violence.
A different kind of sophistication arrived with the defector hunts of the 1950s, the sophistication of disguise. The cyanide spray device that Stashynsky used against Rebet and Bandera was engineered to do something the ice axe could not: to kill while concealing that a killing had occurred. The aim was a death that a coroner would record as a heart attack. This is the methodology of the deniable elimination, and it represents a real technical advance, the product of the poison laboratory and of a doctrine that valued, above all, the absence of evidence. The Markov pellet of 1978 belongs to the same family. The ricin sphere fired from the umbrella was a marvel of miniaturized engineering, designed so that the victim would feel only a momentary jab in a crowd and would, ideally, die days later of a mysterious illness with no obvious cause. The defining ambition of this entire middle period was concealment. The perfect operation was one the world could not even recognize as an operation.
Polonium broke the pattern in 2006, and the break is the single most important observation in this section. Polonium-210 is, in the narrow sense, a sophisticated weapon, because it can be produced only by a state with nuclear infrastructure. But it is the opposite of a concealing weapon. It leaves an indelible radioactive trail that turned the investigation of Litvinenko’s death into a map of his killers’ every movement. If the goal in 1957 had been a death no one could attribute, the polonium case produced a death that could be attributed almost immediately and was, within a decade, formally attributed by a judicial inquiry to the apparatus of the Russian state. The Salisbury attack of 2018 pushes the same logic further. A military-grade nerve agent deployed on a suburban door handle is not a precision instrument of quiet elimination. It is conspicuous, it is indiscriminate, as the death of Dawn Sturgess proved, and it is unmistakably the product of a state weapons program.
A useful way to see the shift is to ask what a coroner would write. In the ideal cases of the middle period, the coroner would write heart failure, and the file would close with the truth unrecorded. In the modern cases, the coroner’s findings became the opening chapter of an international incident, naming a radiological isotope or a battlefield nerve agent and pointing, by the very rarity of the substance, toward a state. The forensic signature that the older doctrine worked so hard to erase is, in the newer operations, almost flaunted. No service that genuinely wished to leave no trace would reach for a material that can be manufactured in only a handful of installations on earth. The persistence of such choices across two separate operations, twelve years apart, against two different kinds of target, is too consistent to be dismissed as coincidence or simple incompetence.
That trajectory, then, is not a simple line from crude to sophisticated. It runs from crude, through a long middle period devoted to concealment, and arrives at something that is technologically advanced and operationally conspicuous at the same time. The modern Russian operation does not hide the hand of the state. In a meaningful sense, it advertises it. A useful comparison is with the program that the Israeli service built, which has long combined real operational secrecy with a quiet, semi-acknowledged tolerance at home, a balance examined in the account of Israel’s decades of targeted killings. The Russian path is different. It has moved toward methods that are, almost by design, traceable. Why a state would deliberately choose traceability is the puzzle, and the answer lies in what the killings are for.
Key Figures
Institutions do not kill. People do, on the orders of other people, and the history of Soviet and Russian assassination is carried by a relatively small number of individuals whose careers illuminate the system that employed them. Four figures, drawn from across the eight decades, stand out.
Pavel Sudoplatov, the Architect
No single person did more to shape the Soviet practice of killing abroad than Pavel Sudoplatov. He began as the operative who personally handed Yevhen Konovalets the box of exploding chocolates in Rotterdam in 1938, the operation that established the disguised-weapon technique. He rose to organize Operation Utka, the campaign that killed Trotsky, designing the layered, two-network structure that allowed the patient approach to succeed after the crude one failed. Through the war and into the postwar years he ran the specialized unit responsible for what the service called wet affairs, overseeing the laboratory that developed poisons and the planning of operations against the emigration. Sudoplatov’s career embodies the institutionalization of assassination. He was not a rogue or a fanatic; he was a senior professional who treated the elimination of state enemies as a technical problem to be solved with planning, patience, and the right tools. He fell with the leadership reshuffles that followed Stalin’s death, spent years in prison, and in old age published memoirs that, for all their self-justification, confirmed in detail what the service had done. He is the human link between the Konovalets bomb and the Trotsky axe, and through the unit he built, between those operations and everything that came after. To read his account is to understand that the program was never the work of monsters in the popular sense. It was the work of competent administrators who had been taught that a particular kind of murder was simply part of the job, and who carried that teaching forward into the institutions that trained the next generation.
Ramon Mercader, the Assassin
Ramon Mercader represents a different figure in the system: not the planner but the instrument, the penetration agent who becomes the weapon. His significance lies in what his career demonstrates about access. The Trotsky operation did not succeed because the axe was a clever weapon. It succeeded because Mercader had spent years constructing a false identity and a false relationship that placed him, unsuspected, behind his victim’s chair. He absorbed a twenty-year prison sentence without breaking his cover, never publicly admitting who had sent him, and was rewarded only in the secrecy of the Soviet honors system. Mercader is the model of the disciplined human asset, and his discipline is the counterpoint to the later failure of Bohdan Stashynsky, the assassin who did break, who defected and confessed, and whose confession taught Moscow how fragile the deniable killing really is. The contrast between the two men, the one who kept silent for twenty years and the one who spoke, frames the central operational vulnerability of any assassination program: it depends on people, and people are not reliably silent. Every planner who came after had to weigh that lesson, and the bloc model of the Markov killing, the practice of lending lethal expertise to junior services rather than always deploying one’s own hands, was in part an institutional response to it.
Andrei Lugovoi and Dmitry Kovtun
The two men who shared tea with Alexander Litvinenko at the Millennium Hotel belong together as a single entry, because their case illustrates the modern relationship between the state and the people it uses. Andrei Lugovoi was a former KGB and FSB officer who had moved, like many of his generation, into the private business world of post-Soviet Russia while retaining the contacts and the trust of the security establishment. Dmitry Kovtun was his associate. The British inquiry found that the two men carried the polonium that killed Litvinenko and that they were acting on behalf of others. What makes their case instructive is the aftermath. Far from being disowned, Lugovoi was protected. Russia refused every request for his extradition, and he was elected to the Russian parliament, where parliamentary status carried legal immunity. The trajectory from suspected assassin to legislator is its own kind of statement. It tells every potential operative that the state stands behind its people, and it tells the watching world that accountability, in this system, runs in only one direction.
The Officers of Salisbury
The two GRU officers who traveled to Salisbury in 2018, operating under the false identities Petrov and Boshirov, complete the gallery. They matter less as individuals than as evidence of a shift. The older operations in this history were carried out, where possible, by cut-outs and proxies, by Spaniards and Bulgarians and recruited foreigners, by people whose connection to Moscow could be denied. The Salisbury operatives were serving officers of a Russian state intelligence service, and when their cover was stripped away by investigators and journalists, what stood revealed was not a deniable proxy but the state itself. Their televised insistence that they were tourists admiring a cathedral spire was so implausible that it functioned, in effect, as a non-denial, a denial no one was expected to believe. They are the figures in whom the modern doctrine of conspicuous killing becomes visible in human form: state employees, barely concealed, performing a cover story as ritual rather than as genuine concealment. Where Mercader’s entire value lay in never being traced to his masters, the Salisbury officers’ eventual exposure did not, in the end, derail anything that Moscow appeared to care about. That difference, between an operative whose worth depended on invisibility and operatives whose visibility cost the state very little, measures the distance the doctrine has traveled.
Consequences and Impact
The cumulative effect of eight decades of operations cannot be measured only in the lives ended, although that toll, counting the named cases and the many lesser-known ones, is substantial. The deeper impact lies in three areas: the chilling of the Russian emigration, the steady damage to Russia’s relations with the states on whose soil it has acted, and the precedent the program has set for the wider world.
A first consequence is the one the operations were most directly designed to produce. Every defector, every exiled critic, every former officer who broke with the service has had to live with the knowledge that the reach of the apparatus does not stop at the border. The killing of Trotsky told the emigration of the 1940s that even the most famous and best-guarded enemy could be reached. The defector hunts told the intelligence officers of the 1950s that crossing over did not buy safety. The Litvinenko and Salisbury cases told the modern wave of Kremlin critics living in London and elsewhere that the calculation had not changed. This is the intended impact, the manufacture of fear as a tool of control over a population that has physically escaped Russian jurisdiction but cannot escape Russian attention. It is, in the cold logic of the program, its principal product. The effect is cumulative across generations, because each new operation refreshes the lesson, and the lesson is always the same: exile is not the same thing as safety, and a new name on a new passport is not a guarantee of being left alone.
Damage to Russia’s external relationships forms a second consequence, and here the ledger is more ambiguous than the program’s designers might wish. The Litvinenko killing drove British-Russian relations to their lowest point since the Cold War, produced the expulsion of diplomats, and left a judicial finding of state responsibility on the public record. The Salisbury attack produced something larger: a coordinated expulsion of Russian intelligence officers from numerous Western countries acting in concert, one of the broadest such responses ever mounted. Each operation, in other words, carried a real diplomatic cost. Yet none of these costs proved durable enough to halt the program. Diplomats returned, relations were partially repaired, and the suspects remained beyond the reach of the courts that wanted them. The pattern suggests that Moscow has consistently judged the diplomatic price worth paying, which is itself a significant fact about how the Russian state values the deterrent effect of the killings against the international friction they generate. There is a harder edge to this calculation that deserves to be named. A state that is willing to absorb diplomatic punishment for its killings has, in effect, discovered that the international system has no mechanism strong enough to stop it. Expulsions are reversible. Sanctions can be endured. A judicial inquiry in a foreign capital produces a document, not an arrest. Each time Moscow weathered the response and kept its operatives safe, it learned, and taught every observer, that the practice is sustainable.
A third consequence is the one that reaches beyond Russia. The Soviet and Russian record is the longest-running demonstration in modern history that a major power can kill its enemies on foreign soil and absorb the consequences. Every state that has since developed or expanded a program of extraterritorial elimination has operated in a world where that demonstration already existed. The American drone campaign, examined in the comparison of the drone program in Pakistan, the Israeli campaigns, and the contested record of eliminations attributed to Indian agencies all unfold against a backdrop in which the Russian precedent is a known quantity. The legal and normative arguments about whether and when a state may kill abroad, set out in the survey of the international law of targeted killing, have been shaped in part by the long Soviet record of treating sovereignty as terrain. Moscow’s program did not create the practice of state assassination, which is older than any modern intelligence service. But its persistence, its scale, and its survival across the collapse of one political system and the rise of another have made it the reference case against which every other program is, consciously or not, measured.
Analytical Debate: Sophistication, Recklessness, and the Logic of Being Seen
A central interpretive question raised by this history is the one the methodology section set up. The Russian program has moved, over its modern decades, toward weapons that are conspicuous and traceable rather than hidden. Two readings of that shift compete, and the choice between them matters for how the wider world should understand the threat.
One reading holds that the modern operations represent a decline, a loss of the tradecraft that the service once prized. On this view, the Soviet apparatus at its peak valued the deniable killing, the cyanide spray engineered to mimic a heart attack, the ricin pellet designed to leave a mystery rather than a crime. The Litvinenko polonium trail and the Salisbury nerve agent that killed an uninvolved bystander look, by that older standard, like failures: operations that exposed the state, generated judicial findings of responsibility, and triggered mass diplomatic expulsions. The proponents of this reading point to the farcical cover story of the Salisbury officers and the radioactive trail across London as evidence of a service that has lost its discipline, that recruits and deploys less carefully than it once did, and that has confused the capacity to kill with the capacity to kill well. Recklessness, on this account, is exactly the right word.
A competing reading holds that the shift is not decline but design, and that to call it recklessness is to misunderstand the purpose. On this view, the older preference for concealment belonged to a specific situation: the Stashynsky affair had shown that an exposed killing could become a public indictment, and for a period the service grew genuinely cautious. But the modern Russian state, this reading argues, has made a different calculation. The point of killing a defector is not merely to remove that individual. It is to discipline the entire population of potential defectors and critics, and that disciplinary effect depends on the killing being understood as the work of the state. A death that is successfully disguised as a heart attack deters no one, because no one learns a lesson from a heart attack. A death by polonium, documented over three weeks of public dying and then formally attributed to the Russian leadership by a foreign court, teaches a lesson with maximum clarity. The conspicuousness is not a bug. It is the function. Russia, on this reading, kills to be seen killing, and the diplomatic costs it absorbs are simply the price of the message.
Weighing the two readings, the honest position is that neither is complete and that both capture part of the truth. It is entirely possible for a single operation to be both a deliberate demonstration and a botched one. The choice of polonium and of a nerve agent may well reflect an intentional preference for methods that signal state capability, while the radioactive trail, the surviving Skripals, and above all the death of Dawn Sturgess reflect genuine operational failures within that intentional choice. The deepest point may be this: a doctrine that values being seen has, by its own logic, lowered the premium it once placed on precision. When concealment is no longer the goal, the discipline that concealment demanded erodes, and the result is operations that are conspicuous by intent and clumsy by consequence. The named disagreement, sophistication or recklessness, is in the end a false binary. The modern Russian operation is recklessness in the service of a message, and the message is the part that is not an accident.
There is one further consideration that should temper any confident verdict, and it concerns the limits of the evidence itself. The operations that historians and journalists are able to analyze in detail are, by definition, the ones that were exposed. The cyanide-spray killings of Rebet and Bandera were revealed only because Stashynsky defected and confessed; without that defection, both deaths would still stand in the record as natural. It is therefore entirely plausible that the modern era contains successful, concealed eliminations that have never been recognized as the work of a state at all, and that the conspicuous cases such as the polonium and Novichok attacks are not a representative sample of Russian practice but a visible minority of it. This caveat does not overturn the analysis. The shift toward conspicuousness is real and documented in the cases that can be seen. But whether conspicuousness is the whole story of what Moscow has been doing, or only the part that came to light, is a question the public record, by its very nature, cannot fully answer. Any honest account of this history has to hold that uncertainty in view.
Why It Still Matters
This history is not a closed file. It is the background to a present in which the practice of states reaching across borders to kill is, if anything, more widespread and more openly discussed than at any point since the Cold War. Three reasons explain why the long Soviet and Russian record continues to repay close study.
A first reason is that it establishes the baseline for the deterrence calculation that every state with such a program must make. Moscow’s experience demonstrates, across eight decades, that a determined major power can sustain a program of extraterritorial killing through diplomatic crises, judicial findings of responsibility, and coordinated expulsions, and emerge with the program intact. That is a sobering lesson for anyone who hopes that international censure alone will constrain such behavior. It is also a lesson that the designers of newer programs will have absorbed. When analysts examine the modus operandi of contemporary campaigns, including the patterns dissected in the study of covert eliminations and their tradecraft, they are looking at practices that operate in a world the Soviet program helped to normalize.
Beyond deterrence, a second reason is that the Russian case clarifies, by contrast, the distinctive features of other programs. The comparison with the Israeli service is instructive. Both states maintain long-running capabilities for killing enemies abroad, but they have made opposite choices about visibility and accountability. The Israeli campaigns that grew from the Munich massacre, and the later operations against Iranian nuclear scientists examined in the account of that sabotage and assassination effort, have generally been conducted with real operational secrecy and a measure of domestic legal scrutiny. The Russian program has moved the other way, toward conspicuous methods and toward an aftermath in which the state protects rather than disowns its operatives. India’s contested record, set against this spectrum, occupies its own position, and the value of placing all three side by side is that it reveals how differently states can answer the same question. The choice is never simply whether to kill abroad. It is how, how visibly, and with what relationship between the act and the state that ordered it.
A related point follows from the comparison, and it concerns the question of accountability. The Russian record is, among other things, a long demonstration of what happens when a powerful state faces no domestic check on this kind of operation. The defector hunts, the umbrella, the polonium, the nerve agent: in none of these cases did a Russian court, a Russian parliament, or a Russian press hold the operation to account, and the suspects who were identified abroad were sheltered rather than surrendered. The contrast with states whose programs have been subjected to at least some legislative or judicial scrutiny is stark, and it suggests that the character of an extraterritorial killing program is shaped less by the technology available to it than by the political system that governs it. A program answerable to no one tends, over time, toward exactly the conspicuous impunity that the modern Russian cases display. That is a lesson with relevance well beyond Russia, for any state that builds such a capability and then has to decide how, and whether, to constrain it.
A third reason is the human one, and it is easily lost in the strategic analysis. Behind every operation in this history is a person who was alive and then was not, and frequently a wider circle of harm that the planners did not intend and did not care to prevent. Trotsky, an old man at a desk. Markov, a writer walking to a bus stop. Litvinenko, dying in a hospital bed over three weeks while his organs failed. And Dawn Sturgess, who had nothing to do with any of it, killed by a weapon of war discarded in a public place. The strategic vocabulary of deterrence and signaling and tradecraft can make the killing sound like a chess problem. The Salisbury case, with its innocent dead, is the standing refutation of that comfort. It is the reason this history matters not only to intelligence analysts but to anyone who lives in a city where the long reach of a hostile state might one day be exercised. The doctrine that began with an ice axe in Coyoacan has not been retired. It has only changed its weapons.
Frequently Asked Questions
Q: What is the history of KGB assassination operations?
The practice of killing enemies abroad was built into the Soviet security services from their founding as the Cheka in 1917, and it continued through every successor agency, the OGPU, the NKVD, the MGB, the KGB, and after 1991 the FSB, the SVR, and the military intelligence service the GRU. The earliest operations targeted exiled opponents of the Bolshevik Revolution, including the kidnapping of White Russian generals in Paris in 1930 and 1937. The killing escalated under Stalin, with the murders of defecting intelligence officers and nationalist leaders in the 1930s, reached a peak of ambition with the killing of Leon Trotsky in Mexico in 1940, and continued through the Cold War with operations against Ukrainian nationalists and defectors. After the Soviet collapse the doctrine survived, producing the polonium poisoning of Alexander Litvinenko in London in 2006 and the Novichok nerve-agent attack in Salisbury in 2018. The continuity across eight decades and multiple changes of name is the defining feature of the record.
Q: How did the KGB kill Trotsky?
Leon Trotsky was killed in August 1940 at his fortified home in the Coyoacan borough of Mexico City, in an operation codenamed Utka, the Russian word for duck, organized by the senior officer Pavel Sudoplatov. The operation used two independent networks. The first, led by the muralist David Alfaro Siqueiros, staged a pre-dawn raid in May 1940 in which around twenty men fired hundreds of rounds into the house; Trotsky and his wife survived by lying flat against a wall. The second network relied on a penetration agent, the Spaniard Ramon Mercader, who had spent years building a false identity and a trusted relationship within Trotsky’s circle. On August 20, 1940, Mercader struck Trotsky on the head with a shortened mountaineering ice axe while the revolutionary read a manuscript. Trotsky died the next day. Mercader served twenty years in a Mexican prison without admitting his connection to Soviet intelligence and was later secretly decorated as a Hero of the Soviet Union.
Q: What was the poison umbrella murder?
The poison umbrella murder refers to the killing of the Bulgarian dissident writer Georgi Markov in London in September 1978. Markov, a defector who broadcast criticism of the Bulgarian Communist regime for the BBC and Radio Free Europe, was crossing Waterloo Bridge when he felt a sharp jab in his thigh from a man carrying an umbrella. He fell ill that evening and died four days later, on September 11, 1978. A post-mortem recovered a tiny metal pellet, smaller than a pinhead, drilled with cavities that had carried a microscopic dose of the poison ricin. Investigators concluded the umbrella had been modified into a device to fire the pellet into a target in a crowd. A nearly identical pellet was recovered from another Bulgarian defector, Vladimir Kostov, who had been attacked in Paris and survived. The political decision was made in Sofia, but the toxin and the pellet technology were supplied by the Soviet service, which had the laboratory expertise the Bulgarian organs lacked.
Q: How was Litvinenko killed with polonium?
Alexander Litvinenko, a former officer of the Russian Federal Security Service who had defected to Britain and become a sharp public critic of the Russian state, was poisoned on November 1, 2006, when he drank tea at the Pine Bar of the Millennium Hotel in London with two Russian men, Andrei Lugovoi and Dmitry Kovtun. The tea contained polonium-210, a rare and intensely radioactive isotope produced in nuclear reactors. A quantity smaller than a grain of salt is lethal once swallowed. Litvinenko fell ill within hours, his condition deteriorated over three weeks as his organs failed, and he died on November 23, 2006, the first known victim of deliberate polonium poisoning. A British public inquiry reported in January 2016 that Lugovoi and Kovtun had carried out the poisoning, that they had probably been directed by the FSB, and that the operation had probably been approved at the highest levels of the Russian state. In 2021 the European Court of Human Rights found Russia responsible for the killing.
Q: How have Russian assassination methods evolved over time?
The methods have followed an unusual trajectory. The earliest operations were crude and physical: a bomb disguised as chocolates in 1938, a hail of submachine-gun fire and then an ice axe against Trotsky in 1940. The Cold War defector hunts of the 1950s introduced disguise as the priority, above all the cyanide spray device used by the assassin Bohdan Stashynsky, which was engineered to make murder look like a natural heart attack. The Markov pellet of 1978 belonged to the same family of concealing weapons. Then the pattern broke. The polonium used against Litvinenko in 2006 and the Novichok nerve agent used in Salisbury in 2018 are technologically advanced but conspicuous, leaving trails that lead straight back to the state. The trajectory runs from crude, through a long middle period devoted to concealment, to a modern phase of methods that are advanced and traceable at the same time.
Q: Does Russia kill to be seen killing?
There is a strong argument that it does, at least in the modern period. The purpose of killing a defector or an exiled critic is not only to remove that one person but to discipline the entire population of potential defectors and critics. That disciplinary effect depends on the killing being understood as the work of the state. A death successfully disguised as natural deters no one, because no one draws a lesson from it. A death by polonium, documented across three weeks of public dying and then formally attributed to the Russian leadership by a foreign judicial inquiry, communicates a lesson with maximum clarity. On this reading, the conspicuousness of modern Russian operations is not a failure of tradecraft but a deliberate feature, and the diplomatic costs Moscow absorbs are the price of the message. The competing view is that the conspicuousness simply reflects declining professionalism. The most defensible position is that both are partly true: the methods are chosen to signal state power, and that very choice erodes the discipline that quiet killing once required.
Q: How does the KGB tradition compare to Mossad’s targeted killings?
Both Russia and Israel maintain long-running capabilities for killing enemies abroad, but they have made opposite choices on visibility and accountability. The Israeli program, which grew most visibly out of the campaign to hunt the perpetrators of the 1972 Munich massacre, has generally been conducted with genuine operational secrecy and has been subject to a degree of domestic legal scrutiny, including a notable Israeli Supreme Court ruling on the lawfulness of targeted killing. The Russian program has moved in the opposite direction, toward conspicuous methods and toward an aftermath in which the state shields rather than disowns its operatives, as when a Litvinenko suspect was later elected to the Russian parliament. In short, the Israeli model leans toward secret act and semi-acknowledged policy, while the modern Russian model leans toward conspicuous act and officially denied but barely concealed responsibility.
Q: Is Russia’s assassination program getting more reckless?
By the standard the Soviet service once set for itself, yes. The older doctrine prized the deniable killing, the operation the world could not even recognize as an operation. Measured against that standard, the polonium trail that mapped Litvinenko’s killers across London, the survival of both Skripals in Salisbury, the transparently false tourist cover story of the Salisbury officers, and above all the death of the uninvolved bystander Dawn Sturgess all look like serious failures. Whether recklessness is the complete explanation is contested. An alternative view holds that the modern program has deliberately abandoned concealment because being seen is now part of the purpose, and that what looks like recklessness is partly the predictable erosion of precision once concealment is no longer the goal. Either way, the modern operations are markedly more conspicuous and have produced more collateral harm than the Cold War norm.
Q: What is the difference between the KGB, the FSB, the SVR, and the GRU?
The KGB was the unified Soviet security and intelligence agency from 1954 until the Soviet collapse in 1991, handling both domestic security and foreign intelligence. When the Soviet Union dissolved, the KGB was broken up. Its domestic security functions passed to the FSB, the Federal Security Service. Its foreign intelligence directorate became the SVR, the Foreign Intelligence Service. The GRU is separate from all of these: it is the military intelligence service, attached to the armed forces, and it has operated more or less continuously since the Russian Civil War, never part of the KGB. In the history of assassination, the distinction matters because different operations are attributed to different services. The Litvinenko killing was attributed to people linked to the FSB, while the Salisbury attack was attributed to officers of the GRU. The phrase KGB tradition is used as shorthand for the institutional culture of state killing that all of these services inherited.
Q: Who was Pavel Sudoplatov?
Pavel Sudoplatov was the single most important figure in the Soviet history of organized assassination. He personally carried out the 1938 killing of the Ukrainian nationalist leader Yevhen Konovalets in Rotterdam, handing him a box of chocolates that was in fact a bomb. He then organized Operation Utka, the campaign that killed Trotsky in 1940, designing its layered two-network structure. Through the Second World War and into the postwar years he ran the specialized unit responsible for what the service called wet affairs, the bureaucratic euphemism for assassination, and oversaw the laboratory that developed poisons and disguised weapons. Sudoplatov was a senior professional rather than a fanatic, a man who treated the killing of state enemies as a technical problem. He fell from favor in the reshuffles after Stalin’s death and spent years imprisoned, later publishing memoirs that confirmed much of what the service had done.
Q: Why was polonium-210 chosen to kill Litvinenko?
This is one of the most revealing questions in the whole case, because polonium is in many ways a strange choice. It is not a quiet poison. It is produced in nuclear reactors, which points investigators toward a state as its source, and it leaves an indelible radioactive trail that, in Litvinenko’s case, allowed scientists to map his killers’ movements across more than forty locations. If the only goal had been a discreet, untraceable death, polonium would have been a poor selection. The likely reasons it was chosen lie elsewhere. It killed slowly and visibly, over three weeks of widely reported public dying. It required resources only a state could marshal, which itself communicated capability. And it delivered a message of impunity that an ordinary weapon could not. The polonium operation looks, in retrospect, less like a concealed elimination than like a demonstration.
Q: Did the KGB really use exploding chocolates and other disguised weapons?
Yes. The disguised weapon, the device built to look like a harmless everyday object, was a recurring feature of Soviet tradecraft. The 1938 killing of Yevhen Konovalets used a bomb concealed in a box of chocolates. The Cold War assassin Bohdan Stashynsky used a device that sprayed crushed cyanide vapor to simulate a heart attack. The 1978 killing of Georgi Markov used a pellet the size of a pinhead, drilled to carry ricin and fired from an umbrella modified into a pneumatic weapon. The service maintained a dedicated laboratory, known informally as the Kamera, for developing such poisons and devices. The thread connecting these cases is a sustained institutional investment in weapons designed either to be undetectable or to be delivered without the victim realizing an attack had occurred.
Q: What happened to Dawn Sturgess in the Salisbury case?
Dawn Sturgess was an English woman with no connection to espionage, to Russia, or to the intended target Sergei Skripal. Several months after the March 2018 nerve-agent attack in Salisbury, her partner found what he believed to be a discarded bottle of perfume in the area. Investigators concluded it was in fact the counterfeit perfume bottle that the attackers had used to carry the Novichok nerve agent and had discarded after the operation. Sturgess sprayed the contents on her wrists, fell gravely ill, and died on July 8, 2018. Her death is the clearest single piece of evidence of the danger of the method chosen. A military-grade nerve agent left in a public place killed an entirely innocent person who happened to find it, and her death stands as the irreducible refutation of any claim that the operation was precise or professional.
Q: Were the Salisbury attackers ever brought to justice?
No. British investigators identified two officers of the GRU, Russian military intelligence, who had traveled to Salisbury under the false names Alexander Petrov and Ruslan Boshirov. The two men appeared on Russian television and denied involvement, claiming they had simply been tourists who came to see Salisbury Cathedral and its spire, an account that was widely ridiculed. Investigative journalists subsequently published what they presented as the officers’ real identities and intelligence backgrounds. Russia, however, has consistently denied state responsibility for the attack and has not handed over the suspects. The same pattern of non-extradition followed the Litvinenko case, in which Russia refused to surrender the named suspects and one of them went on to a career in Russian politics with parliamentary immunity. The practical outcome in both cases is that no one has faced trial.
Q: How did the world respond to the Litvinenko and Salisbury attacks?
Each attack carried a real diplomatic cost. The Litvinenko killing drove British-Russian relations to their lowest point since the Cold War, led to the expulsion of Russian diplomats, and produced a judicial inquiry that placed a finding of probable state responsibility on the public record. The Salisbury attack triggered a larger reaction: a coordinated expulsion of Russian intelligence officers from numerous Western countries acting together, one of the broadest such responses ever mounted, alongside the independent confirmation of the nerve agent by the international chemical-weapons watchdog. Yet neither response proved durable enough to halt the underlying program. Diplomats eventually returned, relations were partially repaired, and the suspects remained beyond the reach of the courts. The pattern indicates that Moscow has repeatedly judged the diplomatic price worth paying.
Q: What was the Stashynsky affair and why did it matter?
Bohdan Stashynsky was a Soviet assassin who killed two Ukrainian nationalist figures in Munich, Lev Rebet in 1957 and Stepan Bandera in 1959, using a device that sprayed cyanide vapor to disguise the deaths as heart attacks. For a time both operations appeared successful. Then, in 1961, Stashynsky, troubled by what he had done and influenced by his wife, defected to the West. He confessed in detail, described the spray weapon, and named his handlers. He was tried publicly in West Germany in 1962, and the court treated him in important respects as a tool of the state that had directed him, placing the heavier moral and legal weight on the Soviet leadership. The affair was a serious setback for Moscow because it converted two deniable killings into a public, documented indictment of the Soviet state. It taught the service that the deniable killing is only deniable until the assassin decides to talk, and for a period it made the leadership more cautious about close-quarter operations against exiles.
Q: How does the Russian record compare to other countries’ targeted killing programs?
The Soviet and Russian record is the longest continuous case study of state assassination available, predating the modern Israeli, American, and Indian programs. Its distinctive features stand out by contrast. Where the Israeli program has generally combined operational secrecy with a degree of acknowledged policy and domestic legal review, the modern Russian program has favored conspicuous methods and official denial paired with the visible protection of its operatives. Where the American drone campaign operated through a legal framework, however contested, and through a remote technology, the Russian approach has remained close-quarter and personal, built around poisons and disguised weapons. India’s contested record occupies its own position on this spectrum. The value of the comparison is that it shows how differently states answer the same questions, not only whether to kill abroad but how visibly, and with what relationship between the act and the state behind it.
Q: Why does the history of KGB assassinations still matter today?
It matters for three reasons. First, it establishes the baseline for deterrence: the Russian record demonstrates, across eight decades, that a major power can sustain a program of killing abroad through diplomatic crises, judicial findings, and mass expulsions and still keep the program intact, which is a sobering lesson for anyone relying on international censure to constrain such behavior. Second, it clarifies by contrast the choices made by other states, throwing into relief the different balances they strike between secrecy, accountability, and visibility. Third, and most importantly, it is a human story rather than a strategic abstraction. Behind every operation is a person who was alive and then was not, and sometimes a wider circle of unintended harm, as the death of the bystander Dawn Sturgess in Salisbury made unbearably clear. The doctrine that began with an ice axe in 1940 has not been retired. It has only changed its weapons, and that is why the record continues to demand attention.