On a September morning in 1968, a slim, soft-spoken police officer named Rameshwar Nath Kao walked into a borrowed set of rooms in New Delhi with twelve hundred staff, no charter that the public would ever read, and an instruction from the Prime Minister that amounted to a single sentence: build India an external intelligence service that will never again let the country be surprised. Kao had watched two wars expose the limits of the old arrangement. He had seen the Intelligence Bureau, designed to watch India from within, fail to read the Chinese build-up before 1962 and stumble again before 1965. The new organisation he was asked to create would look outward instead, and it would carry a deliberately bland name chosen so that nobody outside government would think about it twice.

That organisation was the Research and Analysis Wing, and the story of the next fifty-eight years is the story of how a defensive intelligence service, conceived to prevent surprises, allegedly became an offensive instrument capable of reaching into Pakistani cities and removing India’s most wanted men one by one. The distance between Kao’s founding charter and the campaign that analysts now call the shadow war is the distance between gathering information and ending lives, and the agency crossed that distance gradually, era by era, failure by failure. Nobody at the founding meeting in 1968 imagined motorcycle assassins in Karachi or coded references to unknown gunmen in Pakistani police reports. Yet the institution that produces such operations today is a direct descendant of the small, careful service Kao assembled, and understanding the shadow war means understanding the half-century of evolution that made it possible.

This is the institutional biography of that service. It traces the organisation through five distinct eras, names the men who led it, examines the operations that defined it and the failures that reshaped it, and confronts the central question that hangs over its modern reputation: did the external service evolve naturally toward offensive operations, or did it abandon the defensive mandate its founder believed in? The answer matters, because the campaign of targeted eliminations now unfolding across Pakistan is, more than anything else, the product of how this particular institution learned, decade after decade, what it could and could not achieve by other means.

The World That Produced RAW

To understand why India built a dedicated external intelligence service in 1968, you have to understand the wreckage of the years immediately before it. The country had emerged from Partition with an intelligence apparatus inherited almost entirely from the colonial state. The Intelligence Bureau, founded in the nineteenth century to monitor the independence movement and other internal threats, was the only professional intelligence body the new republic possessed. It was competent at watching India. It was never built to watch the world.

Consider what the new republic actually inherited at independence. The British had built an intelligence apparatus optimised for the surveillance of Indian nationalists, the monitoring of communist activity, and the management of frontier tribes. It was an instrument of imperial control turned inward, and when the British departed in 1947, the Indian state simply took it over and kept it pointed in the same direction. The Intelligence Bureau watched dissent, tracked political movements, and reported on internal stability. Foreign intelligence, in any systematic sense, was a residual function performed by a small number of officers without dedicated resources, doctrine, or training. A country surrounded by a hostile Pakistan and an increasingly assertive China had, in effect, no eyes pointed outward.

The first catastrophe arrived in October 1962. China’s offensive across the Himalayan frontier found the Indian Army unprepared, outnumbered in key sectors, and operating on assumptions that turned out to be badly wrong. Political leadership in New Delhi had convinced itself that Beijing would not fight. The intelligence picture that should have corrected that assumption did not exist, because the body responsible for foreign intelligence was a domestic security bureau working far outside its competence. B.N. Mullik, the long-serving Intelligence Bureau director, carried much of the blame, fairly or not, for a failure that humiliated the country and broke Jawaharlal Nehru’s health and confidence in his final years.

What made 1962 so devastating was not merely the military defeat but the completeness of the surprise. Indian planners had not understood the scale of the Chinese build-up, had not read Beijing’s intentions, and had not given the political leadership anything resembling a usable warning. A modern state had walked into a war it did not see coming, and the institution that should have prevented that blindness was structurally incapable of the task. Nehru’s faith in Chinese friendship, the policy summarised in the slogan of Hindi-Chini brotherhood, had no corrective check from a competent foreign intelligence service, because no such service existed.

Three years later came the second shock. The 1965 war with Pakistan was not the disaster that 1962 had been, but it exposed the same structural weakness. Indian planners were caught off guard by the scale of Pakistan’s infiltration into Kashmir during Operation Gibraltar, and the intelligence flowing to decision-makers was patchy, slow, and divided among bodies that did not coordinate. A country that had now fought two wars in three years, both preceded by intelligence breakdowns, could no longer pretend the problem would solve itself.

There was also a bureaucratic dimension to the lesson of 1965, one that mattered as much as the strategic one. The intelligence failures of the two wars were not only failures of collection. They were failures of organisation. A single bureau, designed for a domestic mission and answerable to the Home Ministry, was being asked to also run foreign espionage, technical collection, and strategic analysis, and it could not give the foreign task the focus, the specialist staffing, or the political attention it required. The problem was structural, and structural problems require structural solutions. The realisation that India needed a separate, purpose-built foreign service, rather than simply more resources for the existing bureau, was the genuine intellectual breakthrough of the mid-1960s. It was a recognition that domestic and foreign intelligence are different crafts, demanding different skills, different cultures, and different lines of accountability. That recognition is the true origin point of the Research and Analysis Wing, and it explains why the new service was deliberately built outside the Home Ministry and close to the Prime Minister, where the foreign mission could not be diluted by the noise of domestic policing.

Indira Gandhi, who became Prime Minister in January 1966, understood the lesson with unusual clarity. She had inherited a strategic environment in which India faced a hostile, nuclear-curious China to the north and a Pakistan that had twice gone to war and was actively rebuilding for a third round. She needed an organisation that could see those threats coming. Crucially, she also wanted an instrument that answered to the Prime Minister’s office rather than to the sprawling Home Ministry, an arrangement that would keep external intelligence close to the centre of national power and away from bureaucratic dilution.

The man she turned to was Rameshwar Nath Kao. Kao was not a flamboyant figure. He was a Kashmiri Pandit, an Indian Police Service officer of quiet manners and precise habits, a connoisseur of art who collected sculpture and spoke softly enough that subordinates leaned in to hear him. He had run the Aviation Research Centre and had handled sensitive assignments, including the investigation into the 1955 Kashmir Princess bombing, the attempt on Zhou Enlai’s life that killed sixteen people aboard an Air India aircraft. He had earned a reputation as a careful operator who understood the craft and never sought the spotlight. Mullik had marked him as a coming man. Indira Gandhi marked him as the founder of her new service.

Those who worked alongside Kao in the founding years left a consistent portrait. He was courteous to the point of formality, deeply read, fluent in the cultural registers of the countries he worked against, and almost pathologically discreet. He believed that an intelligence officer’s first duty was self-effacement, that the best operations were the ones nobody ever heard about, and that an external service should be invisible not only to its enemies but to its own public. He cultivated foreign counterparts with the patience of a diplomat and the wariness of a professional. Officers who later rose to lead the agency described him less as a boss than as a mentor whose standards they spent their careers trying to meet.

Kao’s mandate, as he understood it, was specific and defensive. The Research and Analysis Wing would collect foreign intelligence, run human sources abroad, develop technical collection, and analyse what it gathered so that Indian leaders would never again be blind to a gathering threat. It would focus on Pakistan and China above all. It would build liaison relationships with friendly services. It would, in Kao’s conception, be a thinking organisation, an instrument of foresight rather than force. The CIA provided one model of how a modern external service could be structured, and Kao studied foreign services closely, but he was building something shaped to India’s particular needs and constraints rather than a copy of anyone else.

It is worth dwelling on what that founding mandate actually emphasised, because the contrast with the present is so stark. Kao’s priority list was dominated by anticipation. He wanted advance warning of Chinese military movements. He wanted insight into Pakistani strategic intentions. He wanted a service that could tell the Prime Minister what was coming before it arrived, so that the humiliations of 1962 and the surprises of 1965 would never repeat. The instrument was conceived as a shield, a way of buying the political leadership time and clarity. The idea of the service as a sword, an instrument that would itself strike at named enemies abroad, was simply not part of the founding conversation.

What the founding charter did not contain, in any form the founder would have recognised, was a licence to kill. Kao believed in covert action where it served the national interest, and the operations of his first decade would prove that he was willing to use the instrument boldly. But covert action in Kao’s understanding meant shaping events, supporting friendly forces, gathering decisive intelligence, and occasionally engineering political outcomes. It did not mean a standing campaign of assassination on foreign soil. That idea would have struck the founding generation as both operationally reckless and strategically counterproductive. The institution that now allegedly runs such a campaign had to travel a long way from the room where it began, and the journey was neither quick nor planned.

The Rise

The history of the external service divides cleanly into five eras, each defined by a dominant strategic problem, a characteristic posture, and the leadership that shaped it. The Founding Era ran from 1968 to 1977 and was defined by Kao himself. The Expansion Era ran from 1977 to roughly 1990, the years of Sri Lankan entanglement and the Sikkim achievement. The Retrenchment Era covered the 1990s, when political support thinned and the Kashmir insurgency consumed the agency’s attention. The Modernisation Era ran from the turn of the century through the trauma of the Mumbai attacks and the structural reforms that followed. The Offensive Era began in the middle of the last decade and produced the campaign that defines the institution’s modern reputation. Walking through these eras, and through the men who led the service during each, is the clearest way to see how an organisation built to prevent surprises learned to deliver them.

The Founding Era and the Kao Decade

Kao led the Research and Analysis Wing for its first nine years, and those years established almost everything about the institution’s character. He recruited a generation of officers who would later be called the Kaoboys, a label that mixed affection with a hint of mystique. Many came from the police services, some from the military, a few from academia and the foreign service. Kao looked for officers who could think rather than merely follow procedure, and he built a culture that prized analytical judgement, linguistic skill, and the patience to run sources over years rather than weeks.

The structure Kao assembled in those early years was modest by the standards of the great powers but ambitious for a developing state. He built desks organised by geography and by function. He invested in the Aviation Research Centre, which gave India an aerial reconnaissance capability aimed primarily at the Chinese frontier. He developed early technical collection against communications targets. He helped shape the Special Frontier Force, a covert paramilitary unit raised after 1962 and drawn substantially from the Tibetan exile community, an instrument that could operate in the high Himalayan terrain where conventional forces struggled. None of this was the work of a single year, but the architecture of a serious external service took shape with surprising speed once the political backing was secure.

What distinguished the founding generation was less the hardware than the human tradecraft. Kao believed that the decisive advantage in intelligence came from people, from sources patiently cultivated, from officers who understood the languages and the politics and the personalities of the countries they worked against. He sent young officers abroad under cover, taught them to think in decades rather than news cycles, and insisted that analysis was as important as collection. A piece of raw information, in Kao’s view, was worthless until a thinking officer had placed it in context and drawn out its meaning. That conviction, that the agency was fundamentally an instrument of understanding, is the deepest layer of the founding inheritance.

The defining achievement of the founding decade arrived almost immediately. When East Pakistan erupted in 1971, as the Pakistani military launched a brutal campaign against the Bengali population, Kao’s young service was handed the opportunity that would justify its existence. The agency trained, armed, and helped direct the Mukti Bahini, the Bengali resistance, feeding intelligence to the Indian military and helping prepare the ground for the campaign that ended in the creation of Bangladesh in December 1971. It was the most consequential covert operation in Indian history, and it is examined in full in our reconstruction of the 1971 liberation campaign. For an institution barely three years old, Bangladesh was both a triumph and a template. It proved that a well-run external service could change the map of South Asia.

How the 1971 operation actually worked is instructive. The agency did not simply hand weapons to the Bengali resistance and hope for the best. It helped organise the Mukti Bahini into something resembling a coherent force, established training arrangements, built intelligence networks inside East Pakistan that fed the eventual conventional campaign, and coordinated with the Indian military so that the covert and the conventional tracks reinforced each other. When the formal war came in December 1971, the Pakistani military in the east was already exhausted, surrounded by a hostile population, and starved of the local intelligence advantage that an occupying force needs. The surrender of more than ninety thousand Pakistani troops in Dhaka was a conventional military outcome, but the conditions that made it possible had been shaped, over months, by intelligence work.

The second great achievement of the Kao decade was Sikkim. The Himalayan kingdom’s absorption into the Indian Union in 1975 was, in significant part, an intelligence-driven outcome. The external service cultivated political forces inside Sikkim, shaped the conditions under which the monarchy lost legitimacy, and helped engineer the referendum and political process that ended with Sikkim becoming an Indian state. Critics then and since have called it regime change by quieter means. Admirers have called it the cleanest covert success the agency ever achieved. Either way, it confirmed that the institution Kao built could deliver strategic results without firing a conventional shot.

Sikkim deserves attention because of how cleanly it was done. There was no insurgency to arm, no proxy army to raise, no risk of an autonomous armed actor turning against its sponsor. The operation worked through politics, through influence, through the patient cultivation of local actors whose own interests aligned with the outcome New Delhi wanted. The Chogyal, Sikkim’s monarch, found his position eroding under a combination of domestic political pressure and Indian strategic interest until the kingdom voted itself into the Union. For students of the agency’s later evolution, Sikkim is the purest example of what Kao believed covert action should be: decisive, deniable in its mechanics, achieved through the shaping of events rather than the shedding of blood.

Yet even in this founding decade, the limits of Kao’s defensive conception were visible. The agency built an aviation reconnaissance capability, ran technical collection against China, and developed liaison ties with several foreign services. It was bold. It was creative. It was also, fundamentally, an instrument that shaped events through proxies, intelligence, and political engineering. It did not maintain a standing programme of lethal operations against named individuals abroad. The founder’s service was an architect of outcomes, not an executioner of people, and that distinction would matter enormously when later generations began to blur it.

The Founding Era ended abruptly in 1977. Indira Gandhi’s defeat at the polls, following the Emergency she had imposed, brought the Janata government of Morarji Desai to power, and Desai distrusted the external service intensely. He associated it with the abuses of the Emergency and with the Prime Minister who had created it. Kao left. K. Sankaran Nair, his deputy and one of the original architects of the service, took the chair briefly, but the new political climate was hostile. Budgets were cut. Operations were curtailed. Desai’s well-known indiscretion about a sensitive intelligence channel, an episode that intelligence veterans still discuss with anger, did real damage to a source relationship. The institution that had helped create a nation in 1971 spent the late 1970s fighting for its own survival inside the Indian state.

This first crisis taught the institution a lesson about its own fragility that it never forgot. An external service that depends entirely on the favour of whoever holds power is an external service permanently one election away from being dismantled. The Janata interlude demonstrated that the agency’s reach, its budget, and even its operational secrets could be put at risk by a change of government. Officers who lived through those years drew a hard conclusion: the institution needed to embed itself deeply enough in the machinery of the state that no single political swing could gut it. The drive toward institutional permanence, toward becoming a fixture rather than a favour, dates from this period of vulnerability.

The Expansion Era and the Sri Lankan Wound

Indira Gandhi returned to power in 1980, and with her return the external service regained political backing. Kao came back into government as a senior security adviser, his counsel sought even though he no longer held the chair. The agency entered what can fairly be called its Expansion Era, a period in which it took on a wider range of operations across the neighbourhood and, in doing so, learned a hard lesson about the cost of covert action that goes wrong.

The defining episode of this era was Sri Lanka. In the early 1980s, as Tamil grievances on the island hardened into armed insurgency, the external service trained and supported several Tamil militant groups, including, in those early years, the organisation that would become the Liberation Tigers of Tamil Eelam. The logic at the time was a familiar one in the covert-action tradition: a friendly proxy could pressure Colombo, protect Tamil interests, and give New Delhi leverage over a neighbour drifting toward unwelcome alignments. For a few years the policy seemed to work.

It then turned into the worst entanglement in the agency’s history. The Indian Peace Keeping Force deployed to Sri Lanka in 1987 found itself fighting the very Tigers that Indian intelligence had earlier helped to arm and train. The mission bled for three years before withdrawing in 1990 without victory. The wound did not close with the withdrawal. In May 1991, the LTTE assassinated former Prime Minister Rajiv Gandhi at an election rally in Tamil Nadu, killing a man who had ordered the peacekeeping deployment and turning the entire episode into a national trauma. For the external service, Sri Lanka became the cautionary tale that every later generation would invoke: a proxy you build can outgrow you, turn on you, and exact a price you never imagined.

Sri Lanka was not the only theatre of the Expansion Era. Through the 1980s the agency confronted the Khalistan insurgency, which drew sustenance from Pakistani support and from networks among the Sikh diaspora. The external service developed counter-terror cells, the specialised units later known by designations such as CIT-X and CIT-J, to work against Khalistani and Kashmir-focused networks and, by some accounts, to carry pressure into Pakistan itself. This counter-pressure tradition, contested and never officially confirmed, is part of the institutional inheritance examined in our analysis of India’s covert operations doctrine. The agency of the 1980s was active, willing to take risks, and increasingly drawn into the grey zone between intelligence and covert force. But it was also an agency repeatedly burned by the unintended consequences of that grey zone, and the burns left scar tissue that shaped the caution of the decade that followed.

The Khalistan period is important for a reason that goes beyond the insurgency itself. It was during these years that the agency, by the accounts of several retired officers, first developed a sustained capability for counter-pressure operations against Pakistan. The logic was reciprocal. If Pakistan’s intelligence establishment was funding and arming separatists who killed Indians, then India would develop the means to raise the cost of that policy inside Pakistan. The phrase associated with this period, that a Mumbai answered by a Karachi, captured a doctrine of tit-for-tat covert pressure. How far that doctrine was actually implemented remains contested, and India has never confirmed it, but the institutional memory of having developed such a capability in the 1980s and 1990s is part of the lineage that runs forward to the present campaign.

The Khalistan insurgency itself was eventually contained, primarily through a hard, controversial, and effective police campaign inside Punjab during the early 1990s, combined with the drying-up of popular support and the disruption of the cross-border supply networks. The external service’s contribution lay in mapping and degrading the international dimension of the movement, the diaspora financing, the foreign sanctuaries, the Pakistani handlers. That experience, the experience of dismantling a terrorist network by attacking its external scaffolding rather than only its frontline fighters, became a template the institution would draw on decades later. The separatist movements of the present, and the diaspora networks that sustain them, are confronted with methods whose origins lie in the Khalistan campaign of the 1980s and 1990s.

A succession of chiefs led the service through these years, professionals drawn largely from the police and the original Kaoboy generation, men such as G.C. Saxena and A.K. Verma who carried Kao’s institutional culture forward while managing an organisation under growing operational strain. The leadership of the Expansion Era inherited Kao’s emphasis on analysis and human intelligence, applied it to a harder set of problems, and handed on to their successors an institution that had learned, painfully, that covert action could wound the hand that used it.

The Retrenchment Era and the Kashmir Decade

The 1990s were, for the external service, a decade of difficult problems and thin political reward. The dominant strategic emergency was the Kashmir insurgency, which exploded after 1989 as Pakistan’s intelligence establishment poured trained fighters, weapons, and money across the Line of Control. The external service found itself stretched, its attention pulled toward the Valley, its resources consumed by an internal conflict that was being fed from across the border.

This was the era in which the rivalry between India’s external service and Pakistan’s Inter-Services Intelligence took on the bitter, decades-long character analysed in our study of the intelligence war between the two agencies. For most of the 1990s, the contest was asymmetric in Pakistan’s favour. The ISI ran proxy war as deliberate state policy, infiltrating fighters, managing militant organisations such as Lashkar-e-Taiba, and treating cross-border terrorism as a low-cost instrument against a larger neighbour. India’s external service, by contrast, was still primarily a collector and analyser of intelligence. It could see the proxy war clearly. It could not, within the doctrine of the time, respond in kind.

The texture of the Kashmir decade is worth recalling, because it explains the frustration that built inside the institution. Through the 1990s, militant groups crossed the Line of Control in steady waves. Training camps operated openly on the Pakistani side. Weapons, money, and trained fighters flowed across a frontier that India could monitor but not seal. The external service produced report after report documenting the architecture of Pakistani support, the camps, the handlers, the financing, the safe houses. Those reports were accurate. They were also, in a sense, useless, because the doctrine of the time offered no answer beyond diplomacy and the deployment of more troops to the Valley. The agency was watching a war it was not permitted to fight.

Political support for the agency fluctuated through the decade as governments rose and fell in New Delhi. Coalition politics made intelligence a contested resource, and the institution lacked the steady backing it had enjoyed in the Kao years. Officers of the period have written about a service that was professionally serious but strategically constrained, fully aware of the threat building across the border yet operating under a doctrine that limited it to defence.

The decade ended with the failure that would force the first wave of serious reform. In the summer of 1999, Pakistani forces and irregulars occupied heights on the Indian side of the Line of Control in the Kargil sector. The intrusion took New Delhi by surprise. Indian troops fought a hard, high-altitude campaign to evict the occupiers, and they won, but the strategic surprise was a humiliation for the intelligence community. A.S. Dulat, who led the external service around this period and later wrote with unusual candour about his years handling Kashmir, presided over an organisation that was forced to confront how badly it had read the build-up. The Kargil Review Committee that followed examined the entire intelligence architecture and concluded that India’s agencies did not share information well, did not coordinate, and did not give decision-makers a coherent picture. The Retrenchment Era ended with a verdict: the defensive model, as practised, was not even reliably succeeding at defence.

Kargil cut deep precisely because it was a surprise of exactly the kind the agency had been founded in 1968 to prevent. The whole rationale of an external intelligence service was anticipation, and here, three decades after the founding, India had again been caught unaware on its own frontier. The pattern that had produced the trauma of 1962 had repeated. For an institution whose founding myth was the prevention of strategic surprise, the Kargil intrusion was not merely an operational failure. It was an existential indictment, and the reforms that followed were driven by a determination that the founding promise would finally be made real.

The Modernisation Era and the Mumbai Reckoning

The turn of the century opened the Modernisation Era. The Kargil failure produced the first structural overhaul of India’s intelligence machinery in decades. New bodies were created to fill the gaps the review committee had identified. A dedicated technical intelligence organisation was established to consolidate signals and technical collection. A defence intelligence agency was set up to serve the military. The post of National Security Adviser, created in 1998, became the apex coordinator of the intelligence community, and the external service increasingly worked within that coordinated architecture rather than as a free-standing instrument reporting only to the Prime Minister.

These reforms changed the institution in ways that were not always visible from outside. The creation of a national technical research organisation meant that signals intelligence, satellite imagery, and other technical disciplines could be developed at scale and shared across agencies. The strengthening of the National Security Adviser’s office meant that the external service’s product was integrated into a single national assessment rather than competing with the reporting of rival bodies. The agency was, slowly, being woven into a coordinated machine. For an organisation that had spent the 1990s producing accurate reports that no one acted on, the prospect of being part of a system that could actually convert intelligence into decision was genuinely transformative.

Vikram Sood, who led the agency in the early years of the decade and later became one of the most articulate public commentators on Indian intelligence, presided over a service adjusting to this new environment. The Modernisation Era was, in its first years, genuinely a story of reform: better technical collection, clearer coordination, a more professional relationship with the military and the policy establishment. The agency was becoming a more capable collector and analyser of intelligence.

What it was not yet becoming was an offensive instrument. Even after Kargil, even with the proxy war from Pakistan running at full intensity, the doctrine held. The external service modernised its tools and its structures while leaving its fundamental posture intact. India would gather intelligence, build international pressure, and pursue Pakistan-sponsored terrorism through diplomacy and law. The kinetic option remained, in official doctrine, off the table.

There was a logic to that restraint, and it deserves to be stated fairly. India in the early 2000s was a rising economic power that wanted integration into the global system, foreign investment, and a reputation as the responsible democracy of South Asia. An assassination programme on foreign soil sat badly with that self-image. Pakistan, in this period, was a frontline state in the American campaign in Afghanistan, which gave it diplomatic cover and made any Indian covert pressure inside Pakistan diplomatically expensive. The defensive doctrine was not merely institutional timidity. It was a considered judgement that India’s interests were better served by patience, growth, and the slow accumulation of international goodwill than by a covert war whose costs were unpredictable.

What it was not yet becoming was an offensive instrument. Even after Kargil, even with the proxy war from Pakistan running at full intensity, the doctrine held. The external service modernised its tools and its structures while leaving its fundamental posture intact. India would gather intelligence, build international pressure, and pursue Pakistan-sponsored terrorism through diplomacy and law. The kinetic option remained, in official doctrine, off the table.

Then came the night of 26 November 2008. Ten heavily armed men from Lashkar-e-Taiba, dispatched by sea from Karachi, came ashore in Mumbai and began a three-day assault that killed one hundred and sixty-six people across hotels, a railway station, a hospital, and a Jewish centre. The full reconstruction of that siege is set out in our complete account of the Mumbai attacks, but its meaning for the external service can be stated briefly. There had been intelligence. There had been warnings about a possible sea-borne attack on Mumbai. The intelligence had not been sharp enough, had not been acted upon decisively, and had not prevented the catastrophe. For a community that had spent the years since Kargil reforming itself precisely so this would not happen, the Mumbai siege was a verdict on the limits of the defensive model.

What made Mumbai so corrosive to the institution’s self-image was the granular nature of the failure. This had not been a complete intelligence vacuum, as 1962 had been. Warnings had existed. Technical intelligence had picked up indications of a sea-borne threat. The problem was that the system, even after a decade of post-Kargil reform, had still not converted those fragments into a sharp, actionable, believed warning delivered to the people who could have hardened Mumbai’s defences. The reformed machine had still failed. That was a far more disturbing conclusion than a simple absence of information, because it suggested that the defensive model might be structurally incapable of reliability no matter how much it was reformed.

The reforms that followed were the most far-reaching since the founding. A National Investigation Agency was created to pursue terrorism cases across state lines. A national intelligence grid was designed to connect databases. Coastal security was overhauled. Coordination mechanisms were strengthened again. The chiefs who led the agency in the years after Mumbai, professionals such as Sanjeev Tripathi and Alok Joshi, managed an institution under intense pressure to demonstrate that it had learned.

But the deeper lesson of Mumbai was not bureaucratic. It was strategic, and it was corrosive. The siege demonstrated, with horrifying clarity, that the entire apparatus of defence, the intelligence reforms, the coordination bodies, the diplomatic pressure on Pakistan, could not reliably stop a determined, state-enabled attack. India responded to 26/11 with restraint, choosing diplomacy and international isolation of Pakistan over military retaliation. That restraint won India sympathy. It did not win India a single conviction in Pakistan of the men who planned the slaughter. The masterminds remained free. The organisation that dispatched the killers continued to operate. Within the institution and within the wider strategic community, a question began to harden: if defence cannot stop the attacks, and if reactive engagement cannot deliver justice, what is left?

Equally damaging was the long aftermath. In the months and years after Mumbai, India built a meticulous case. It shared dossiers of evidence with Pakistan. It provided intercepts, the confession of the surviving attacker, and the testimony that connected the plot to handlers on Pakistani soil. It pressed through every diplomatic and legal channel available. And it watched, year after year, as the men it had named walked free, as prosecutions in Pakistan stalled, as the organisation responsible rebranded and continued. The lesson the institution drew was brutal in its simplicity. The defensive and legal toolkit, deployed with maximum diligence against the clearest possible case, had produced no justice at all. That recognition, more than any single reform, is what eventually broke the doctrine of restraint.

But the deeper lesson of Mumbai was not bureaucratic. It was strategic, and it was corrosive. The siege demonstrated, with horrifying clarity, that the entire apparatus of defence, the intelligence reforms, the coordination bodies, the diplomatic pressure on Pakistan, could not reliably stop a determined, state-enabled attack. India responded to 26/11 with restraint, choosing diplomacy and international isolation of Pakistan over military retaliation. That restraint won India sympathy. It did not win India a single conviction in Pakistan of the men who planned the slaughter. The masterminds remained free. The organisation that dispatched the killers continued to operate. Within the institution and within the wider strategic community, a question began to harden: if defence cannot stop the attacks, and if reactive engagement cannot deliver justice, what is left?

The Offensive Era

The answer that emerged over the following decade is the subject of the institution’s modern reputation, and it is the reason this history exists. The Offensive Era began, by most assessments, in the middle of the last decade, and it represents the sharpest doctrinal break in the agency’s fifty-eight-year life.

Its first visible signs were conventional and military rather than covert. After an attack on an Indian Army camp at Uri in 2016, India publicly announced cross-border strikes against launch pads on the Pakistani side of the Line of Control. After the Pulwama bombing in 2019, Indian aircraft struck a target at Balakot inside Pakistan proper. These were military operations, openly acknowledged, designed to signal that India would now impose costs across the border. They were not the work of the external service alone. But they marked the abandonment of the post-26/11 restraint, and they created the political space in which a quieter, deadlier track could develop.

The significance of the 2016 and 2019 strikes lay less in their military effect than in what they did to the doctrine. For decades, the unwritten rule had been that India absorbs attacks and responds through diplomacy. The Uri strikes broke that rule openly, and the Balakot strike broke it more dramatically still, sending aircraft across the international border for the first time since 1971. Once the principle of crossing the border to impose costs had been publicly established and politically rewarded, the logic that had kept covert lethal operations off the table weakened considerably. If India would openly bomb a target inside Pakistan, the argument for refusing to quietly remove a wanted terrorist there became much harder to sustain.

That quieter track is what analysts now call the shadow war, examined in full in our overview of the campaign of targeted eliminations. Beginning around 2022, a pattern emerged inside Pakistan that had no precedent. Men on India’s most-wanted list, militant commanders and organisers who had spent years living openly under the protection of the Pakistani state, began to die. Some were shot at close range by gunmen on motorcycles. Some died in their homes, in mosques, in markets. The Pakistani authorities, in case after case, could neither prevent the killings nor produce the killers. The recurring signature of these operations is decoded in our study of the unknown gunmen pattern, and the cumulative scale of the campaign is set out in our complete record of the eliminations.

What distinguishes the shadow war from a simple campaign of violence is its evident discipline. The targets are not random. They are, by the consistent pattern of open-source reporting, men with names, files, and histories of involvement in attacks against India. The operations show a recurring method, an understanding of the targets’ patterns of life, and a professionalism in execution and escape that points away from ordinary criminality. The men dying in Pakistani cities are, overwhelmingly, the men an Indian most-wanted list would identify, and they are dying in a manner that suggests careful preparation rather than chaotic happenstance. That pattern is the strongest circumstantial argument that the killings are the product of an organised campaign rather than Pakistan’s ambient violence.

India has never confirmed responsibility for a single one of these deaths. Official spokesmen have dismissed the killings as the result of Pakistan’s own internal violence, of factional disputes within militant organisations, of the lawlessness that Pakistan’s territory generates on its own. That denial is consistent, deliberate, and, for an intelligence operation, entirely expected. Deniability is the point. The strongest external evidence pointing toward Indian involvement came from outside the subcontinent altogether. A major British newspaper published an investigation alleging that Indian intelligence had been behind a series of killings on Pakistani soil. Separately, Canada publicly accused Indian agents of involvement in the killing of a Sikh separatist on Canadian soil, and prosecutors in the United States alleged an Indian-linked plot against a separatist there. These were allegations, contested by New Delhi, but together they shifted the international conversation. The shadow war stopped being purely a Pakistani complaint and became a subject of Western legal and diplomatic concern.

The Western dimension of the allegations changed the strategic calculus considerably. A campaign confined to Pakistani soil could be managed as a bilateral dispute between two adversaries who already expected the worst of each other. Allegations of operations in Canada and the United States were a different matter, because they implicated India in lethal activity on the territory of partners whose friendship India actively cultivates. The contrast in how India handled the two sets of allegations is itself revealing. Toward Pakistan, denial has been flat and dismissive. Toward the Western governments, India’s response has been more careful, more diplomatically textured, in some instances accompanied by promises of inquiry. The institution that runs the shadow war, in other words, recognises that the campaign carries different costs in different theatres, and it calibrates accordingly.

There is a structural reason the offensive turn took the particular form it did, and it returns to the lesson of Sri Lanka. The shadow war is, by every credible account, built on small, deniable, often mercenary cells rather than on a standing armed proxy of the kind the agency once created among the Tamil militants. This is a deliberate design. A proxy army is visible, autonomous, and prone to developing its own agenda, as the Tigers did with catastrophic consequences. A deniable cell, by contrast, is small, disposable, and tightly controlled, and when an operation is complete the cell can disappear without leaving an institutional footprint. The choice of instrument tells you that the modern agency absorbed its own history with care. It wanted the strategic effect of covert lethal action without the strategic liability of an armed force it could not switch off. The shadow war, in its operational architecture, is the Sri Lankan lesson and the Mumbai lesson fused together: the conviction that the planner must be reached, married to the conviction that he must be reached by an instrument that can never become a Frankenstein. Whether that design holds, whether deniable cells really stay deniable and controllable over a campaign measured in years, is one of the open questions that the institution’s own secrecy makes impossible to answer from outside.

The Offensive Era’s modern leadership presided over this transformation. Samant Goel, who led the agency from 2019 to 2023, headed the service through the years in which the offensive posture took shape. His successor, Ravi Sinha, who took the chair in 2023, has led it through the period of greatest operational tempo and greatest international scrutiny. Neither man has spoken publicly about targeted killings, and neither will. But the institution they have led is, by every external assessment, a fundamentally different organisation from the one Kao founded. It runs technical collection on a scale Kao could not have imagined. It works inside a coordinated national-security architecture. And it stands accused, persuasively if not provably, of conducting a standing campaign of lethal operations on the soil of a nuclear-armed neighbour.

That is the arc of the rise: from a defensive collector built to prevent surprises, through expansion and painful entanglement, through retrenchment and modernisation, to an offensive instrument whose alleged operations would have been unthinkable to its founder. The question of whether that arc represents maturation or corruption is the one this history must finally confront. Before it can, the operations and decisions that defined each turning point deserve closer examination.

Major Actions and Decisions

An institution is defined less by its charter than by the choices it makes when the charter runs out of guidance. Five decisions, spread across the agency’s history, did more than any others to shape what it became. Each was a turning point. Each is best understood not as an isolated operation but as a moment when the institution learned something about itself and about the limits of the doctrine it was operating under.

The Decision to Make Bangladesh

The first defining decision was the choice, in 1971, to commit the young external service fully to the creation of Bangladesh. It was not an obvious decision. India faced enormous risks. The United States was tilting toward Pakistan. China was hostile. A failed intervention could have been catastrophic for a service barely three years old.

Kao’s agency chose to commit anyway, and it chose to commit deeply. The external service did not merely observe the Bengali uprising. It trained the Mukti Bahini, organised its command structures, supplied intelligence for the conventional campaign that followed, and helped shape the political framework of the new state. When Bangladesh achieved independence in December 1971, the institution had its founding triumph.

Why did the decision succeed where so many later covert interventions failed? Several factors aligned. The Bengali resistance was a genuine popular movement with deep local roots, not a synthetic proxy manufactured by a foreign service. The Pakistani military was committing atrocities so severe that the resistance enjoyed near-universal local support and significant international sympathy. India’s covert effort was matched by a conventional military commitment that gave the operation a decisive endpoint rather than an open-ended quagmire. And the strategic objective, the separation of a hostile state’s eastern wing, was concrete and achievable rather than vague. The agency had not improvised a miracle. It had recognised a rare convergence of favourable conditions and committed to it with discipline.

The lesson the agency drew from Bangladesh was double-edged. On one hand, it proved that bold covert action, properly resourced and politically backed, could deliver strategic results of the highest order. On the other, it set a standard that no later operation could match, because the circumstances of 1971, a genuine popular uprising, a Pakistani military committing atrocities on a scale that horrified the world, a friendly population eager for Indian help, were a unique convergence. Every later generation of officers grew up measuring itself against Bangladesh, and the comparison bred both ambition and a quiet frustration that nothing else ever came so cleanly together.

The Decision to Arm Tamil Militants

The second defining decision was the choice, in the early 1980s, to train and arm Tamil militant groups in Sri Lanka. The reasoning was strategic and, on its own terms, coherent. Colombo was drifting toward alignments New Delhi disliked. The Tamil population faced real persecution. A friendly proxy would give India leverage and a humanitarian justification at once.

That decision became the institution’s greatest self-inflicted wound. The proxy outgrew its sponsor. The Liberation Tigers of Tamil Eelam, far from remaining a manageable instrument of Indian policy, became an independent military force with its own absolute objectives. When India deployed peacekeepers in 1987, the agency’s former protégés became the agency’s enemy. The peacekeeping mission failed. And in 1991, the proxy that Indian intelligence had once helped to build assassinated a former Indian Prime Minister.

The lesson of Sri Lanka entered the institution’s DNA. Covert action that creates an autonomous armed actor is a wager against the future, and the future does not always cooperate. The blowback principle, the idea that today’s instrument can become tomorrow’s threat, is examined as a general phenomenon in our broader treatment of covert operations and their consequences. For the external service specifically, Sri Lanka created a deep institutional caution about proxies, a caution that helps explain a striking feature of the later shadow war: the alleged modern campaign relies, by most accounts, on small mercenary cells and deniable contractors rather than on a standing armed proxy. The agency that was burned by the Tigers learned not to build another one.

The Decision Not to Retaliate for Mumbai

The third defining decision was, in a sense, a decision not to act. In the days after the 2008 Mumbai siege, India’s leadership faced enormous pressure to retaliate militarily against Pakistan. The intelligence community, the military, and significant parts of the political class argued for a forceful response. India chose restraint. It pursued diplomatic isolation of Pakistan, built an international case, shared evidence, and pressed for prosecution of the planners.

That restraint was defensible. A military response carried the risk of escalation between two nuclear-armed states. The international sympathy India earned by absorbing the attack without retaliating was real and valuable. But the decision had a long shadow. The planners of the Mumbai attack were not punished. The organisation that carried it out continued to function under state protection, a pattern examined in our analysis of the Pakistani military’s relationship with militant leadership. India’s restraint, in the cold institutional memory of the intelligence community, came to be read as a strategy that protected the attacker more than it served the victim.

The lesson the institution drew from the Mumbai non-retaliation was the most consequential of all. It was the conclusion that the defensive and diplomatic toolkit, the toolkit Kao had built and every later chief had refined, had a ceiling, and that the ceiling was below the level required to actually stop or deter state-enabled terrorism. The shadow war is, more than anything else, the institutional answer to the question that the Mumbai decision left unanswered. It is the doctrine of restraint giving way, over a decade, to a doctrine that imposes a personal, lethal cost on the men who plan attacks.

It is worth being precise about the causal chain, because it is the heart of the institution’s modern self-justification. The argument inside the strategic community runs as follows. Terrorism directed by a state succeeds because the people who plan it are insulated from consequence. The attackers themselves may die in the operation, but the organisers, the recruiters, the financiers, and the handlers live in safety, protected by the sponsoring state, beyond the reach of the victim’s law. If that insulation could be removed, if the planners could be made to feel personally unsafe, then the cost-benefit calculation that makes state-sponsored terrorism attractive would change. The shadow war, on this reasoning, is not revenge. It is deterrence by the only means that the previous decade proved could work, the removal of the planner’s personal safety. Whether one finds that argument persuasive or alarming, it is the genuine institutional logic, and it was forged in the failure of the Mumbai response.

The Decision to Reform After Each Failure

The fourth defining pattern is not a single decision but a recurring institutional behaviour: the agency’s habit of responding to every major failure with structural reform. Kargil produced a wave of new institutions and coordination mechanisms. Mumbai produced another. The external service has spent much of its history rebuilding itself in the aftermath of disaster, and the cumulative effect of those rebuilds is an institution far more technically capable and far better coordinated than the one that existed in 1968.

This habit explains something important about the modern agency. The shadow war is often described as if it were purely a product of political will, of a government deciding to be ruthless. That description is incomplete. The campaign is also a product of capability that was built, reform by reform, over twenty-five years. The technical collection that can locate a target in a Pakistani city, the analytical depth that can build a pattern-of-life file, the coordination that can integrate intelligence with operational planning, all of this is the accumulated residue of post-failure reform. The institution did not simply decide to go on the offensive. It also became, slowly and through repeated trauma, capable of it.

The Decision to Go Offensive

The fifth and final defining decision is the one that created the Offensive Era, and it is the hardest to date precisely, because it was not announced. There was no founding document for the shadow war, no public charter revision, no moment when a Prime Minister stood up and declared that India would now hunt its enemies on foreign soil. The shift happened in the space that intelligence operations are designed to occupy: deniable, gradual, visible only in its results.

What can be said is that the decision crystallised in the years after 2014, took conventional and acknowledged form in the cross-border strikes of 2016 and 2019, and matured into the deniable campaign of targeted eliminations from roughly 2022 onward. By the time international investigators and foreign governments began publicly alleging Indian involvement in killings abroad, the decision was years old and the campaign was well established.

The institutional logic behind the decision is the through-line of this entire history. Bangladesh proved covert action could achieve strategic results. Sri Lanka taught the agency to fear proxies and prefer deniable, controllable instruments. The Mumbai non-retaliation proved the defensive toolkit had a ceiling. The post-failure reforms built the capability. Put those four lessons together, and the fifth decision, the decision to run a standing campaign of lethal operations using small deniable cells rather than armed proxies, looks less like a sudden rupture and more like the conclusion of an argument the institution had been having with itself for fifty years. The shadow war is the answer the external service arrived at after every other answer had been tried and found wanting. Whether it is the right answer is a separate question, and one the institution’s culture is uniquely ill-equipped to debate openly.

There is one further episode that belongs in this account of the institution’s formative decisions, even though it predates the offensive turn by years. In December 1999, the hijacking of Indian Airlines flight IC-814 ended with India releasing imprisoned militants, including a cleric who would go on to found one of the deadliest organisations ever to target the country. The full story of that crisis is told in our account of the IC-814 hijacking. For the external service, the IC-814 episode was a particular kind of wound. It was the experience of being forced, by a hostage crisis it could not resolve operationally, to hand a future enemy his freedom. The men released at Kandahar became, within a few years, architects of attacks that killed Indians. The institutional memory of that forced concession, of the long-term cost of an enemy left alive and free, is part of the psychological background against which the later doctrine of permanent, lethal pressure on named enemies took shape.

The Institution Behind the Operations

Operations and decisions describe what the external service has done. They do not fully explain why it became the particular kind of organisation it is. To understand that, you have to look at the institution itself: how it recruits, how it thinks, what culture it carries, and what structural features shape its behaviour. A psychological portrait of an agency is necessarily an exercise in inference, because the subject is secret by design. But the inferences are not arbitrary. They are drawn from the public writing of retired officers, from the institution’s documented behaviour, and from the consistent shape of its operations over decades.

The first defining feature is the Kao inheritance. The founder built a service that prized analytical judgement, linguistic competence, and the long patience required to run human sources. He recruited thinkers. He valued officers who could read a strategic situation rather than merely process information. That founding emphasis on analysis has never fully left the institution, and it explains why the agency, even in its offensive era, is not a crude organisation. The operations attributed to it in the shadow war are characterised by careful target development, precise timing, and an evident understanding of Pakistani urban geography. That is the Kao inheritance expressed in lethal form: the analytical culture of 1968 applied to the targeting problems of the present.

A second defining feature is the institution’s relationship with secrecy, which is more extreme than that of many peer services. The external service has no public charter that ordinary citizens can read. It is not subject to the kind of parliamentary oversight that scrutinises the CIA or the British services. Its budget is opaque. Its chiefs do not testify in public. This radical secrecy has operational advantages, but it also produces a specific institutional pathology. An organisation that is never required to publicly justify itself is an organisation that struggles to debate its own direction. There is no Indian equivalent of the Church Committee, the public reckoning that forced the American intelligence community to confront its assassination programmes. The external service’s evolution toward offensive operations has therefore happened with almost no public deliberation, and that absence of debate is itself a structural fact about the institution.

The third defining feature is the institution’s memory of its own failures. The agency carries Sri Lanka, Kargil, and Mumbai the way an individual carries formative wounds. These episodes are not abstract history inside the service. They are the reference points against which officers measure risk. The fear of building another LTTE shapes the preference for deniable cells over armed proxies. The memory of Kargil and Mumbai shapes the conviction that the defensive posture has a fatal ceiling. The institution’s offensive turn is, in a real sense, the behaviour of an organisation trying to never feel again what it felt in those years. That is a powerful motivation, and also a dangerous one, because an institution driven by the fear of past failures can talk itself into present recklessness.

A fourth defining feature is the institution’s position inside the Indian state. The external service reports, in effect, to the Prime Minister’s office through the National Security Adviser. It is close to the centre of power and relatively insulated from the wider bureaucracy. This proximity gives the agency reach and political backing, but it also ties the institution’s posture tightly to the strategic preferences of whoever holds power. The shadow war is not the external service acting on its own initiative. It is the external service executing a strategic choice made at the political summit, and the institution’s culture, secretive, offensive-leaning, scarred by failure, makes it a willing and capable instrument of that choice.

A fifth feature, less often discussed, is how the institution recruits and forms the people who carry its culture. In its early decades the external service drew heavily from the police services and the foreign service, with Kao personally selecting officers he judged capable of the patient, ambiguous work of human intelligence. Over time the agency developed its own direct recruitment, drawing analysts, linguists, technical specialists, and area experts into a career structure separate from the rest of the Indian bureaucracy. The formation of an officer is long. Years are spent acquiring a target language, absorbing the political texture of a region, learning the unglamorous craft of agent handling and source verification. This slow apprenticeship has consequences for institutional behaviour. An organisation that takes a decade to fully form an officer cannot turn its culture quickly. It cannot abandon the analytical inheritance even when its missions change, because the inheritance is built into the training pipeline itself. When observers describe the modern agency as offensive, they are describing a change in what the institution does, not a change in the kind of person it produces. The officer running a deniable cell against a target in Pakistan was very likely formed in the same patient tradition as the officer who built the network in East Pakistan in 1970. The continuity of formation is one reason the offensive turn looks, on close inspection, less like a rupture and more like an old institution pointing familiar skills at a new task.

The sixth feature is the technical transformation that has quietly reshaped the service over the past two decades. The agency Kao founded was overwhelmingly a human-intelligence organisation, because in 1968 there was little else it could be. The modern external service operates in an environment saturated with signals, metadata, satellite imagery, financial-tracking systems, and the vast digital exhaust of ordinary life. This has changed the texture of intelligence work. A target who once could be located only through a painstakingly recruited human source can now, in principle, be tracked through the electronic traces he leaves. That technical capability is part of what makes the alleged precision of the shadow war possible. It is far easier to develop a target for elimination when collection is continuous and granular than when it depends entirely on a single agent’s reporting. The technical dimension also explains part of the agency’s growth in size and budget, neither of which is public but both of which are evidently large. The institution behind the operations is therefore not only a culture and a memory. It is also an expanding technical apparatus, and the marriage of that apparatus to the old analytical culture is what produced the capable, dangerous organisation of the present.

The contradiction at the heart of the modern institution sits above all of these features. Kao’s service was built to prevent surprises. The modern agency, in its alleged offensive role, exists in significant part to deliver them. It is an organisation founded on the value of foresight that now allegedly specialises in the unforeseen, the sudden death of a wanted man in a Pakistani street. The institution has not resolved this contradiction. It has simply grown large enough and capable enough to contain both functions at once, the patient analyst and the deniable executioner, inside the same secret structure. That dual personality is the truest portrait of the external service as it exists today.

One last observation about the institution concerns how it sees itself, because self-image shapes behaviour as surely as structure does. Officers of the external service, on the evidence of their published memoirs and their public remarks after retirement, do not regard themselves as members of a rogue agency. They regard themselves as the inheritors of Kao, as professionals carrying a difficult national burden, as the people who stand between the country and the surprises that wrecked it in 1962 and bled it in 2008. This self-image is sincere, and it is part of why the offensive turn has met so little internal resistance. To the officers carrying it out, the shadow war is not a betrayal of the founding tradition. It is the founding tradition, the duty to protect the state, finally equipped with the tools the earlier generations lacked. An outside critic may find that self-understanding troubling, precisely because it is so coherent and so untroubled. An institution that experiences its most controversial activity as simple duty is an institution that will not easily be argued out of it. That, more than any single operation, is the psychological fact that anyone trying to understand the modern external service has to absorb.

Current Status

As of the present, the Research and Analysis Wing stands at the most operationally assertive and the most internationally scrutinised point in its fifty-eight-year history. The institution that Kao founded with twelve hundred staff in borrowed rooms is now a large, technically sophisticated service embedded in a coordinated national-security architecture, and it is, by every external assessment, in the middle of its most active offensive phase.

The leadership reflects the era. Ravi Sinha, who took the chair in 2023, leads a service operating at high tempo. His predecessor, Samant Goel, headed the agency through the years in which the offensive posture matured. Neither chief has publicly acknowledged the campaign of targeted eliminations, and the institution’s culture guarantees that neither ever will. Official India continues to deny involvement in the killings inside Pakistan, attributing them to that country’s internal violence.

The operational picture, as reconstructed from open sources, is one of sustained activity. The pattern of eliminations inside Pakistan that emerged around 2022 has not slowed. If anything, the tempo has increased, and the targets have crept higher up the organisational charts of the militant groups. The institution’s alleged reach now extends, by the accounts of foreign governments, well beyond the subcontinent, into the diaspora politics of Western states. That extension has carried a diplomatic cost. The allegations from a major Western government and the legal proceedings in another have complicated India’s relationships with countries it regards as partners, and they have forced New Delhi into a continuing diplomatic defence of operations it will not admit conducting.

Within the national-security architecture, the external service today occupies a position its founders would scarcely recognise. In Kao’s time the agency was a young organisation fighting for budget, for charter, and for the confidence of a political class that did not yet understand what civilian external intelligence was for. Today it is an established pillar. It sits alongside the domestic intelligence bureau, the military intelligence directorates, the technical-collection agency, and the National Security Council secretariat in a structure that, at least on paper, coordinates the country’s intelligence effort. Coordination among intelligence agencies is never as smooth in practice as it appears in an organisational chart, and the Indian system carries its own rivalries and information silos. But the direction of travel across the past two decades has been toward integration, and the external service has been a principal beneficiary. Its analytical product reaches the political summit. Its operational proposals are weighed at the highest level. The institution that once struggled to be heard now helps set the terms of the conversation.

The debate over oversight has also sharpened in the present period, even if it has produced no change. As the allegations of foreign targeted killings have accumulated, a small but persistent body of Indian commentary has begun to ask whether an institution of this reach should remain entirely outside parliamentary scrutiny. The arguments are familiar from older democracies that travelled this road earlier. Supporters of the status quo say that meaningful oversight of an external service is impossible without compromising sources and methods, and that the agency’s proximity to the Prime Minister’s office already provides political accountability. Critics answer that political accountability exercised entirely in secret is not accountability at all, and that an institution capable of lethal action abroad should answer, in some controlled form, to the legislature. This debate has not moved policy. But its mere existence is a feature of the current moment that would have been unthinkable in Kao’s era, and it suggests that the question of how a democracy governs a powerful secret service has finally arrived, late, in Indian public life.

The institution also operates today against the backdrop of the most serious India-Pakistan military confrontation in decades, the 2025 crisis that the wider analytical community has examined at length. In that environment, the external service functions as one instrument among several, its covert track running alongside conventional military capability in a way that earlier generations of the agency never experienced. The integration of covert and conventional pressure is one of the defining features of the present moment, and it places the external service at the centre of Indian strategy rather than at its quiet periphery.

What has not changed is the secrecy. The agency that the public could not see in 1968 is the agency the public cannot see now. Its budget remains opaque. Its operations remain deniable. Its chiefs remain silent. The institution has become vastly more powerful and vastly more controversial without becoming, in any meaningful sense, more transparent. That combination, rising power and unchanged opacity, is the defining condition of the external service as it exists today, and it is the condition that makes the debate over its evolution so difficult to resolve.

Legacy and Network

The legacy of the external service is not a settled thing. It is a contested inheritance, and the contest is precisely the named disagreement that this history was written to engage: was the agency’s evolution toward offensive operations a natural institutional maturation, or a dangerous departure from the founding mandate? The honest answer is that the evidence supports both readings, and that the choice between them is finally a judgement about values rather than facts.

Begin with the case for maturation, which runs as follows. An intelligence service exists to protect the state. For decades, India’s external service tried to protect the state through the defensive toolkit, intelligence, diplomacy, international pressure, and law, and that toolkit failed, repeatedly and catastrophically, to stop state-enabled terrorism. Kargil happened anyway. Mumbai happened anyway. The planners of mass-casualty attacks lived in safety, protected by a neighbouring state, untouched by the law. In that environment, the argument goes, an institution that refused to develop an offensive capability would have been failing its core duty. The shadow war, on this reading, is the external service finally growing into the full range of capabilities a serious intelligence power requires. It is what Israel’s service did after Munich. It is, in milder form, what the American service did with its drone campaign, a comparison developed in our study of the United States targeted-killing programme. Maturation, in this view, simply means the institution caught up with the world.

The case for dangerous departure runs equally hard in the other direction. Kao did not build an assassination service, and he did not build one because he understood that a standing campaign of lethal operations on the soil of a nuclear-armed rival carries risks that no tactical success can offset. Every elimination is a provocation. Every provocation, between two nuclear states, carries a small but real chance of escalation that no one can fully control. The campaign has also damaged India’s standing with the Western democracies whose partnership it values, transforming a Pakistani grievance into a Canadian and American legal concern. And the campaign has been conducted with no public deliberation whatsoever, by an institution that answers to no meaningful oversight. On this reading, the shadow war is not maturation. It is an unaccountable institution, scarred by failure and freed from scrutiny, talking itself into a permanent low-intensity war whose endgame nobody has defined. The risks of that endgame are taken up in our forward-looking analysis of the future of India’s counter-terror doctrine.

There is a further dimension to the legacy that neither the maturation case nor the departure case fully captures, and it concerns the effect of the external service’s evolution on the strategic environment of the entire region. An intelligence service does not evolve in isolation. When India’s agency shifted toward an offensive posture, it did not simply acquire a new capability. It changed the expectations of every other actor in the South Asian intelligence contest. Pakistan’s own service reads the eliminations on its soil as a declaration, and it calibrates its own behaviour accordingly. The two agencies have always existed in a relationship of mutual adaptation, each shaping the other through decades of action and reaction, a dynamic explored in our comparison of the two rival intelligence services. The offensive turn has therefore not only altered what India’s service does. It has raised the temperature of the contest itself, and a contest run at a higher temperature is a contest with a narrower margin for the misreading, the accident, or the unauthorised act that turns a covert exchange into an open one. The legacy of the offensive era, in this sense, is not just a capability India now owns. It is a regional condition that India now has to manage, and managing it will be the work of the external service for as long as the shadow war continues.

Both cases are serious. Both are honestly held by serious people. What this history can add is not a verdict but a clarification. The evolution of the external service was neither purely natural nor purely chosen. It was, instead, the cumulative product of a half-century of institutional learning, and the lessons the institution learned were real lessons drawn from real failures. The agency did not become offensive because it was corrupt or reckless. It became offensive because, decade after decade, the defensive model produced Kargil and Mumbai and the unpunished planner, and the institution drew the logical, if grave, conclusion. The shadow war is the rational output of the institution’s own painful history. That is exactly what makes it dangerous. A reckless campaign can be corrected by argument. A campaign that is the considered conclusion of fifty years of institutional experience is far harder to stop, because to stop it you have to persuade the institution that everything it learned from its own failures was wrong.

It is worth being honest about what a history like this can and cannot do. It cannot tell a reader whether the shadow war is right. That judgement belongs to citizens, to the political class, and ultimately to the democratic process, such as it is able to reach an institution this secret. What a history can do is insist that the question be asked with the full record in view. The external service did not wake up one morning as an assassination agency. It became what it is through East Pakistan and Sikkim, through the Tigers and the peacekeepers, through Kargil and Mumbai and the long humiliation of the unpunished planner, through reform after reform built on the wreckage of failure after failure. Anyone who wants to praise the offensive turn should know that it grew from genuine catastrophe and genuine institutional learning. Anyone who wants to condemn it should know that it was not recklessness but reasoning, and that reasoning is harder to argue with than recklessness. The most useful thing this account can leave behind is the refusal of the easy version in either direction. The agency is neither the heroic shield of patriotic legend nor the rogue death squad of its harshest critics. It is a serious institution that made a grave choice for serious reasons, and the consequences of that choice are still being written across the region it operates in.

The network the external service has built across fifty-eight years is the infrastructure that will carry this question forward. It is a network of technical collection, of analytical capability, of liaison relationships, of coordinated planning across the national-security architecture, and, allegedly, of deniable operational cells inside hostile territory. That network does not disappear when a chief retires or a government changes. It is the institution’s true legacy, more durable than any single operation, and it means that whatever the shadow war becomes, the capability behind it is now a permanent feature of the South Asian strategic landscape. R.N. Kao built an instrument to prevent surprises. His successors built it into an instrument that can also deliver them. The institution will carry both identities, the analyst and the executioner, for as long as it exists, and the unresolved tension between them is the most important inheritance the founding generation left to the present.

Frequently Asked Questions

When was RAW founded and why?

India’s Research and Analysis Wing was established in September 1968. It was created because India’s intelligence failures before the 1962 war with China and the 1965 war with Pakistan had exposed a structural gap. The Intelligence Bureau was a domestic security organisation, not built to gather and analyse foreign intelligence. Prime Minister Indira Gandhi wanted a dedicated external service that would prevent the country from ever again being strategically surprised by a hostile neighbour.

Who founded RAW and what was his vision?

The founding chief was Rameshwar Nath Kao, a quiet, methodical police officer who had earlier investigated the 1955 Kashmir Princess bombing and run the Aviation Research Centre. His vision was a defensive one. He wanted a thinking organisation that prized analytical judgement, linguistic skill, and patient human-source work, focused above all on Pakistan and China. He believed in bold covert action where it served the national interest, but his conception of covert action meant shaping events and supporting friendly forces, not running a standing campaign of assassinations.

What was RAW’s first major success?

Its first and most consequential success was the agency’s role in the creation of Bangladesh in 1971. The young service trained and helped direct the Mukti Bahini, the Bengali resistance, and supplied intelligence for the Indian military campaign that followed. The independence of Bangladesh in December 1971 was, in significant part, an intelligence-enabled outcome, and it became the founding triumph against which every later operation was measured.

What were RAW’s biggest failures?

Three failures shaped the institution most deeply. The Sri Lankan entanglement of the 1980s, in which the agency armed Tamil militants who later became enemies and whose most powerful faction assassinated a former Indian Prime Minister, was the worst self-inflicted wound. The 1999 Kargil intrusion caught India by strategic surprise. And the 2008 Mumbai siege, despite warnings of a possible sea-borne attack, was a coordination failure that killed one hundred and sixty-six people. Each failure forced a wave of structural reform.

How did RAW change after the Mumbai attacks?

India’s 2008 Mumbai attacks produced the most far-reaching reforms since the founding. A National Investigation Agency was created, a national intelligence grid was designed, coastal security was overhauled, and coordination mechanisms were strengthened. But the deeper change was strategic rather than bureaucratic. Mumbai demonstrated that the defensive and diplomatic toolkit could not reliably stop state-enabled terrorism, and that conclusion began the slow shift toward the offensive posture that defines the institution today.

What is the shadow war and is RAW behind it?

The shadow war is the term analysts use for a pattern of targeted killings inside Pakistan that emerged around 2022, in which militant commanders on India’s most-wanted list have been eliminated by unidentified attackers. India has never confirmed responsibility and officially attributes the deaths to Pakistan’s internal violence. The strongest external evidence pointing toward Indian involvement came from an investigation by a major Western newspaper and from separate allegations by the Canadian and American governments regarding killings and plots on their own soil. The allegations are serious and contested, and deniability is built into the operations by design.

Has RAW’s doctrine really shifted from defence to offence?

Most evidence strongly suggests it has. For much of its history the external service was a collector and analyser of intelligence operating under a defensive doctrine. The cross-border military strikes of 2016 and 2019, openly acknowledged, marked the abandonment of post-Mumbai restraint, and the deniable campaign of targeted eliminations from roughly 2022 onward represents the offensive doctrine in its mature form. Whether this shift is maturation or a dangerous departure is genuinely contested.

How many RAW chiefs have there been?

The agency has had a long succession of chiefs since R.N. Kao led it from 1968 to 1977. Because the institution is so secretive, public knowledge of the full list relies on the accounts of retired officers and media reporting, and the details are not always complete. The best-documented chiefs include Kao himself, his deputy K. Sankaran Nair, A.S. Dulat around the Kargil period, Vikram Sood in the early modernisation years, and the modern chiefs Samant Goel and Ravi Sinha, who have led the service through its offensive era.

Who were the Kaoboys?

The Kaoboys were the generation of officers that R.N. Kao personally recruited and shaped in the agency’s early years. The term mixed affection with a sense of mystique. They were drawn from the police, the military, academia, and the foreign service, and Kao selected them for their ability to think analytically rather than merely follow procedure. The culture they established, prizing judgement, language, and patient source work, has never fully left the institution.

What was RAW’s role in the annexation of Sikkim?

India’s absorption of the Himalayan kingdom of Sikkim into the Indian Union in 1975 was, in significant part, an intelligence-driven outcome. The external service cultivated political forces inside Sikkim and helped shape the conditions and the political process that ended with Sikkim becoming an Indian state. Critics have described it as regime change by quieter means, while admirers regard it as the cleanest covert success the agency ever achieved.

Why did the Sri Lanka operation go so wrong?

In the early 1980s the agency trained and armed Tamil militant groups, including the organisation that became the Liberation Tigers of Tamil Eelam, intending to create a friendly proxy that would give India leverage over Colombo. The proxy outgrew its sponsor. When India deployed peacekeepers in 1987, the former protégés became enemies, the mission failed, and in 1991 the Tigers assassinated former Prime Minister Rajiv Gandhi. The episode taught the institution a lasting fear of armed proxies that helps explain why the modern shadow war reportedly relies on small deniable cells instead.

How does RAW compare to the CIA or Mossad?

All three are external intelligence services, but their reputations differ. The American service is associated with global technical reach and, historically, with a drone-based targeted-killing programme. The Israeli service is associated with a long tradition of targeted operations against named enemies. India’s external service was, for most of its history, a more defensive and analysis-focused organisation. The shadow war has narrowed that gap, drawing comparisons to the Israeli model in particular, though the Indian campaign is more recent and far less publicly acknowledged.

Does RAW have parliamentary oversight?

The external service operates with far less oversight than many peer agencies. It has no public charter that ordinary citizens can read, its budget is opaque, and its chiefs do not testify publicly. There is no Indian equivalent of the public investigations that forced other countries to confront their intelligence practices. This radical secrecy is one reason the institution’s evolution toward offensive operations has happened with almost no public deliberation.

Why does India deny the targeted killings?

Denial is standard practice for deniable intelligence operations everywhere, and it is the entire operational point. If India confirmed responsibility for killings inside Pakistan, it would invite direct retaliation, provide Pakistan with a diplomatic and legal weapon, and risk uncontrolled escalation between two nuclear-armed states. Consistent denial, combined with operations designed to leave no clear attribution, allows India to impose costs on its enemies while limiting the costs imposed in return.

Was RAW’s offensive evolution a natural maturation?

This is the central contested question. One view holds that the defensive toolkit failed so repeatedly to stop state-enabled terrorism that developing an offensive capability was simply the institution growing into the full range of tools a serious intelligence power needs. The opposing view holds that the founder deliberately avoided building an assassination service because of the escalation risks, and that the modern campaign is an unaccountable institution talking itself into a permanent low-intensity war. Both arguments are seriously held, and the disagreement is finally about values rather than facts.

How has the shadow war affected India’s foreign relations?

It has complicated them. As long as the targeted killings were confined to Pakistani soil, they remained largely a Pakistani grievance. But public allegations by Canada regarding a killing on its territory, and a prosecution in the United States alleging an Indian-linked plot, transformed the campaign into a subject of Western legal and diplomatic concern. India has been forced into a continuing diplomatic defence of operations it will not officially admit conducting, straining relationships with countries it regards as partners.

What is RAW’s status today?

The external service is at the most operationally assertive and most internationally scrutinised point in its history. It is led by Ravi Sinha, who took the chair in 2023, and it operates at high tempo within a coordinated national-security architecture, its covert track running alongside conventional military capability. It has become vastly more powerful and more controversial over fifty-eight years without becoming meaningfully more transparent, and that combination of rising power and unchanged secrecy is the defining condition of the institution as it exists now.

How is RAW different from the Intelligence Bureau?

The two organisations divide the Indian intelligence task along a basic line. The Intelligence Bureau is the domestic service, responsible for security threats inside the country, and it predates independence by decades. The external service was carved out in 1968 precisely because the domestic bureau had been overstretched by also trying to run foreign intelligence, a dual burden that contributed to the surprises of the 1962 and 1965 wars. Since then the external service has owned the foreign mission, including the human networks, the analysis of other states, and the covert operations abroad, while the bureau has concentrated on the home front.

How does RAW recruit its officers?

Recruitment has changed across the agency’s lifetime. In the early decades the founder personally selected officers from the police services, the military, the foreign service, and even academia, prizing analytical judgement over procedural experience. Over time the institution developed its own direct recruitment and a distinct career structure, drawing in analysts, linguists, technical specialists, and regional experts. The formation of an officer remains long, often involving years spent acquiring a target language and absorbing the political texture of a region, and that slow apprenticeship is one reason the institution’s analytical culture has proven so durable.

Will the shadow war continue, and what could end it?

Nothing in the present trajectory suggests an early end. The campaign is the considered conclusion of decades of institutional learning, it enjoys political backing at the highest level, and the technical and operational network behind it is now a permanent feature of the regional landscape. The factors that could change that trajectory are external to the agency itself. A serious escalation scare, sustained Western diplomatic pressure, a negotiated shift in the India-Pakistan relationship, or a decision at the political summit to lower the temperature could each alter the calculus. Absent one of those, the offensive posture is likely to persist as the institution’s working doctrine.