Pakistan’s terror infrastructure is not an abstraction. It carries street addresses, satellite coordinates, construction dates, and named administrators. On the night of May 7, 2025, when Indian Rafale jets released SCALP cruise missiles across the international border, the nine sites that absorbed those missiles were not chosen in haste. Indian intelligence had catalogued each location for years, photographed the buildings from orbit, recorded which organisation ran which compound, and assigned every site a serial number. The targets were instruction facilities, and the strike that destroyed them exposed something the public rarely sees laid out in full: the physical network that converts ordinary recruits into operational fighters.

At the briefing it held on the afternoon of May 7, the Indian military displayed a map. That map marked twenty-one facilities spread across Pakistan and Pakistan-occupied Jammu and Kashmir, nine of them ringed as the night’s targets and the remaining twelve catalogued as known but untouched. The display was the clearest public admission of a claim India’s security establishment had pressed for two decades. The neighbouring state does not merely shelter wanted men. It schools them.

This network is the structural target of the wider campaign that has unfolded across the country since 2021. The shadow war that has killed wanted men one by one in Karachi, Lahore, Rawalpindi, and a dozen smaller towns removes graduates. The facilities mapped below produce them. A single elimination subtracts one operative from the rolls. A functioning compound enrols, indoctrinates, and drills a fresh intake every season. Understanding the instruction network, therefore, means understanding why the campaign against individuals has never been treated by Indian planners as sufficient on its own, and why Operation Sindoor escalated from hunting people to bombing the institutions that manufacture them.

A note on what this analysis can and cannot claim is worth making at the outset. The picture of Pakistan’s training network assembled here rests on Indian government briefings, Western intelligence assessments, commercial satellite imagery, and the public record of attacks and strikes. It is, in places, an intelligence picture rather than a courtroom one, built from assessment and inference where ground access is impossible. Where a claim is contested, this account says so. The aim is not to pretend the map is complete or beyond dispute, but to lay out, as honestly as the open record allows, what is known about the facilities that have sat at the centre of the India-Pakistan conflict for thirty years.

The Geography of a National Instruction System

The facilities India catalogued before Operation Sindoor were not scattered randomly across the map. They cluster into three distinct zones, and the logic of that clustering reveals how the system was designed to survive pressure rather than collapse under it.

The first zone is the Punjab heartland. Markaz Taiba sits at Nangal Sahdan in Muridke, Sheikhupura district, on the outskirts of Lahore. Markaz Subhan Allah occupies roughly fifteen acres on the edge of Bahawalpur, three hundred and eighty kilometres to the south. A third Punjab facility, the Sarjal complex at Tehra Kalan, lies in Shakargarh tehsil of Narowal district, around six kilometres from the international boundary near the Samba sector. These heartland sites share a common character. They are large, established, openly built, and located deep inside Pakistan’s most populous and politically dominant province. They function as headquarters, indoctrination campuses, and residential quarters for senior commanders. Their depth inside Punjab is itself a form of protection, because a forward strike against a launch site near the Line of Control does nothing to touch them.

Pakistan-occupied Jammu and Kashmir forms the second zone. Here the density is highest. Around Muzaffarabad, the territory’s capital, intelligence assessments place Sawai Nala, Markaz Syedna Bilal, Chelabandi, Abdullah bin Masood, and Maskar-e-Aqsa. Around Kotli, further south, sit Gulpur, Markaz Abbas, and Maskar Raheel Shaheed. Near Bhimber, closer still to Indian territory, stands Markaz Ahle Hadith Barnala. These sites are forward installations. They are smaller, often disguised inside mosques or seminaries, and positioned to feed recruits across the Line of Control. Distance to that line is the single most important variable describing them. Barnala sits roughly nine kilometres from the Line of Control. Markaz Abbas in Kotli lies about thirteen kilometres from it. The Gulpur facility is around thirty kilometres back. Proximity is not an accident of real estate. It is operational design, because a launch facility loses its value the moment it sits a long march from the frontier.

The third zone runs through the Mansehra belt of Khyber Pakhtunkhwa. Garhi Habibullah, Batrasi, Oghi, Boi, and the camp at Balakot fall within this corridor. The Mansehra district has hosted instruction grounds since the 1980s, when the anti-Soviet jihad in Afghanistan turned the region into a rear base for fighters, weapons, and money. Today the belt provides mid-tier and specialised preparation, set in terrain that is more remote and harder to observe than the open plains of Punjab. The connection to the older Afghan war is not incidental, and it links directly to the tribal-belt landscape that has sheltered armed groups for forty years.

Read together, the three zones form a system with depth. The heartland holds the institutional core. The Kashmir zone holds the forward edge. The Mansehra belt holds a specialised middle layer. A blow against any single zone leaves the other two intact. That redundancy is precisely what makes the network resistant to the elimination campaign, and it is why Indian planners eventually concluded that killing camp graduates one at a time could never match the pace at which the camps replaced them. The geography itself argued for a different kind of strike.

Vocabulary matters when reading this map, because the network’s own terms encode function. A markaz, the word that prefixes so many of the named sites, translates as centre, and it denotes a permanent institutional facility with residential, religious, and administrative roles. A maskar denotes a camp in the narrower military sense, a place built primarily for drilling and physical preparation. A launch pad is something smaller still, a forward staging point near the frontier with no real curriculum, used to brief and equip recruits in the final hours before they cross. A detachment is the most transient category of all, a temporary cluster of tents or rented rooms that can appear and disappear within weeks. Indian intelligence catalogues all four categories, and the public chart of twenty-one sites foregrounds the permanent markaz and maskar facilities because those are the ones that show up consistently on satellite imagery. The launch pads and detachments are more numerous, more mobile, and far harder to fix to a coordinate.

Distance to the Line of Control is not the only spatial logic at work. The network also tracks the country’s road and rail spine. Markaz Taiba sits beside the Grand Trunk Road corridor north of Lahore, with rapid connectivity to the Sialkot border belt and the motorway system. Markaz Subhan Allah at Bahawalpur sits on the southern Punjab rail and road axis. The Mansehra belt is threaded by the Karakoram Highway and the older arteries that once carried supplies to the Afghan front. A facility’s value depends not only on how close it sits to the frontier but on how quickly it can move people and material between the heartland, the specialised middle layer, and the forward edge. This network is, in that sense, a logistics system as much as a set of buildings, and the roads connecting its zones are as much a part of its architecture as the compounds themselves.

The geography of the camps also mirrors the geography of the Pakistani military. The heartland sits in Punjab, the province that supplies the bulk of the army’s officer corps and hosts its general headquarters at Rawalpindi. The Mansehra belt occupies terrain the army has long used for its own mountain warfare schooling. The forward facilities in Pakistan-occupied Kashmir cluster near the same sectors the regular army garrisons, with Markaz Abbas standing roughly two kilometres from the Kotli military camp. This overlap is the spatial signature of state involvement. A network that grew up entirely independent of the state would not so consistently shadow the state’s own footprint. Read carefully, the map of the camps is a faded copy of the map of the army.

Geography is also not static, and the map of 2025 is not the map of 2005. Some facilities have closed, some have moved, and some have been absorbed into the seminary or charity infrastructure around them. A few sites that featured in older intelligence assessments, particularly compounds closest to the Line of Control that drew Indian artillery during periods of tension, were quietly relocated a few kilometres deeper. Others were consolidated when the parent groups rationalised their footprint under financial scrutiny. What the catalogue captures, therefore, is a snapshot of a living system rather than a fixed architecture, and the three-zone structure has held even as individual addresses within it have changed.

Mapping the system finally exposes how cleanly the civilian and the militant overlap. Markaz Taiba contains a functioning mosque, a hospital, and a school. The Sarjal facility at Tehra Kalan is housed inside the building of a Primary Health Centre. Markaz Syedna Bilal in Muzaffarabad shares a three-storey structure with a cupping-therapy clinic on its ground floor. This blending is deliberate. A compound that delivers genuine medical care and genuine schooling cannot be characterised in a single word, and that ambiguity is the network’s first line of legal and diplomatic defence. The dual-use problem runs through every facility on the map, and it is the reason no honest account of this network can pretend the buildings serve only one purpose.

There is a harder edge to that dual-use overlap, and it is worth naming directly. The hospital at Muridke treats real patients. The school at Bahawalpur teaches real children. The clinic on the ground floor of Markaz Syedna Bilal sees real people seeking real treatment. A strike that destroys a militant compound also damages, or risks damaging, the civilian functions braided into it, and the recruits the network loses can be matched by clinic patients and schoolchildren the surrounding population loses. This is among the most difficult facts in the entire subject, and it is exactly why the dual-use design is so effective. It does not merely complicate targeting. It ensures that every option, including doing nothing, carries a human cost, and any honest map of these facilities has to carry that knowledge alongside the coordinates.

Which Organisations Run the Camps

The map India displayed in May 2025 was not a single organisation’s property. Three armed groups operate the bulk of the network, and a fourth actor, the Pakistani state, stands behind all of them.

Lashkar-e-Taiba runs the largest share. Markaz Taiba in Muridke is the group’s institutional core, the facility where the most recruits pass through and where its founding leadership has resided. The Sawai Nala camp at Muzaffarabad, functional since the early 2000s, handles recruitment, registration, and basic preparation. Gulpur near Kotli and Markaz Ahle Hadith Barnala near Bhimber are the group’s forward installations aimed at the Rajouri, Poonch, and Reasi sectors. Lashkar’s reach across all three geographic zones is the reason Indian intelligence treats it as the network’s dominant operator. The organisation is headed by Hafiz Saeed, a man who has cycled through house arrest and conviction without losing his protected status, and whose institutional footprint is examined in detail in the profile of LeT’s headquarters city.

Jaish-e-Mohammed operates the second cluster. Markaz Subhan Allah in Bahawalpur is the group’s operational headquarters, the site Indian briefers connected to the planning of the February 2019 Pulwama attack. Markaz Syedna Bilal in Muzaffarabad is Jaish’s principal facility in the Kashmir territory. Markaz Abbas at Kotli, capable of housing well over a hundred recruits at a time, is run by a founding member of the group. The Sarjal launch complex in Shakargarh is the organisation’s main infiltration point into Jammu and Kashmir. Jaish answers to Masood Azhar, the cleric released by India in the IC-814 hijacking deal of 1999, whose family ties to Bahawalpur are dissected in the account of JeM’s headquarters.

Hizbul Mujahideen holds the third position. The Mehmoona Joya facility near Sialkot is among the group’s largest, and Indian briefers linked it to the planning of the 2016 Pathankot air base assault. Maskar Raheel Shaheed near Kotli, one of the older sites in the Kashmir zone, is assessed as capable of sheltering between a hundred and fifty and two hundred recruits, with specialised instruction in sniping, ambush tactics, and mountain warfare. Hizbul is led by Syed Salahuddin, who also chairs the United Jihad Council, the umbrella body that nominally coordinates several anti-India groups.

The three organisations are not interchangeable, and the differences matter for reading the map. Lashkar-e-Taiba has historically presented itself as a disciplined, mass-membership movement with a strong charitable and missionary face, and its instruction emphasises ideological depth alongside military skill. Jaish-e-Mohammed has historically been smaller, more clerical, more closely identified with a single founding family, and more willing to deploy suicide tactics, the method used at Pulwama. Hizbul Mujahideen carries the strongest claim to a Kashmiri rather than a Punjabi identity and has emphasised recruits who can blend into the population on the Indian side of the line. These doctrinal differences shape what each group’s facilities teach. A Lashkar compound devotes more time to indoctrination and missionary framing. A Jaish facility devotes more attention to the preparation of fidayeen attackers willing to die in place. A Hizbul site such as Maskar Raheel Shaheed leans toward the marksmanship and concealment skills suited to a long campaign inside Kashmir.

The United Jihad Council, which Salahuddin chairs, is the thin coordinating layer above the three groups. It does not run camps of its own. What it provides is a forum through which the separate organisations deconflict their operations, divide recruits, and present a united front to their state patron. The council’s very existence is evidence that the network is managed rather than chaotic, because rival armed groups in genuinely ungoverned conditions tend to fight each other for territory and recruits rather than form umbrella committees.

Behind these three organisations stands the Inter-Services Intelligence directorate and the Pakistan Army. The relationship is not one of passive tolerance. Indian and Western assessments hold that the army supplies serving instructors for weapons instruction at facilities such as Sawai Nala, that it provides transport corridors, and that several of the camps sit beside or inside military real estate. Markaz Abbas at Kotli, by Indian account, lies roughly two kilometres from the Kotli military garrison. The placement is significant, because it removes the most common excuse offered for the network’s existence. A compound operating in an ungoverned tribal periphery can be blamed on a weak state unable to police its own frontier. A compound operating within sight of a regular army cantonment cannot. The role of the intelligence directorate as builder and protector of these groups is the subject of a fuller treatment in the Punjab terror heartland analysis, and it is the single fact that turns a map of buildings into a map of state policy.

The recruits who fill these facilities are not, for the most part, drawn from Pakistan’s educated urban classes. Indian and independent assessments describe an intake weighted toward young men from poor rural districts of Punjab and from the impoverished belts of Sindh and Khyber Pakhtunkhwa, often funnelled through seminary networks that offer food, shelter, and a sense of purpose the state has failed to provide. Smaller organisations feed the same pool. Al-Badr, a group with roots in the older Kashmir jihad, and the residual networks of Harkat-ul-Mujahideen, the organisation several Jaish founders left to build their own outfit, have at various points run or shared facilities within the same geography. The network is dominated by three names, but it has always been broader than three names, and its recruitment base is wide enough to survive the loss of any single organisation.

Pakistan’s intelligence directorate does not run the camps the way an army runs a barracks. Indian and Western assessments describe a handler system, in which serving or retired intelligence officers maintain contact with the groups, set broad direction, supply specialist instruction and equipment, and provide political cover, while leaving day-to-day administration to the organisations themselves. Compartmentalisation is the point of this arrangement. It allows the state to shape the network’s output without leaving fingerprints on any individual facility, and it gives Pakistani officials the deniability they rely on when a compound is exposed. The handler structure is why the relationship is so hard to prove in a courtroom and so hard to deny in an honest assessment, because it is designed to leave influence without leaving evidence.

Formal bans have been imposed on the network before, and the pattern of those episodes is instructive. Pakistan outlawed Lashkar-e-Taiba and Jaish-e-Mohammed in 2002, under intense pressure after attacks that brought India and Pakistan close to war. The organisations did not disappear. They re-emerged under charitable names, their leaders moved between nominal detention and open activity, and the facilities kept operating. Each subsequent round of pressure, the listing of leaders by the United Nations, the financial scrutiny of the grey-list years, has followed the same arc: visible compliance at the surface, continuity underneath. This history is the strongest reason to read any single crackdown sceptically, because the network has spent two decades demonstrating that it outlasts the measures aimed at it.

The state’s involvement also explains the network’s relationship with money and with civilian charity. The parent groups operate vast registered charitable wings: Jamaat-ud-Dawa and the Falah-i-Insaniat Foundation for Lashkar, the Al-Rehmat Trust for Jaish. These bodies run dispensaries, schools, ambulance fleets, and disaster relief, and they collect donations through channels that look entirely legitimate. The charitable apparatus is not a disguise bolted onto the militant apparatus. The two are built together. A donation box that funds a genuine ambulance also funds a compound’s electricity bill, and the seminary that screens orphans for an education also screens them for the first stage of the instruction pipeline. This integration is why the network has proven so resistant to financial pressure, and it is examined more closely in the discussion below of whether the system was ever shut down.

One further detail about the operators matters for anyone trying to track the network through news reports. Under international financial pressure, the parent groups have repeatedly changed names. Lashkar has presented itself as Jamaat-ud-Dawa and later as the Falah-i-Insaniat Foundation, posing as a charity. Newer labels such as The Resistance Front and the People’s Anti-Fascist Front have been attached to attacks in Kashmir to make organised militancy look like spontaneous local resistance, and a further label, Kashmir Tigers, has surfaced in the same role. The buildings did not move when the letterhead changed. A reader following only the name on a press release will lose the thread, while a reader following the coordinates will not.

The Punjab Heartland Facilities

The heartland of the network sits in Pakistani Punjab, and three sites define it. Each deserves a close look, because the heartland facilities are where the system’s scale, its longevity, and its open operation are most visible.

Markaz Taiba at Muridke is the anchor of the entire structure. Established in 2000 at Nangal Sahdan in Sheikhupura district, the complex spreads across a campus that Indian assessments and visiting journalists have described as roughly two hundred acres. Within its boundary wall sit a mosque, a hospital, a school, administrative offices, dormitories, sports grounds, and, by intelligence account, weapons stores and tactical drill areas. Around a thousand recruits are assessed to pass through its courses each year. The site’s notoriety rests on a single fact repeated in every Indian briefing about it: the men who carried out the November 2008 Mumbai assault, including the lone attacker captured alive, were prepared here, and the planners of that operation reportedly visited the campus.

The compound’s history is older than its 2000 construction date suggests. Lashkar-e-Taiba grew out of an organisation founded in the late 1980s by Hafiz Saeed and others to channel volunteers into the Afghan jihad, and Markaz Taiba was built as that organisation pivoted its energy toward Kashmir. Indian and Western reporting has long held that the campus received outside funding in its early years, including a reported contribution from Osama bin Laden toward the construction of a mosque and guesthouse. That detail, whatever its precise accuracy, captures the compound’s place in the wider jihadist economy of the 1990s. Muridke was not a local project. It was a node in a transnational movement, built with transnational money, and designed from the outset to outlast any single campaign.

The compound’s openness is as striking as its history. It appears on commercial satellite imagery in clear detail, it has been visited and photographed by international reporters, and it operated for a quarter of a century without interference from Pakistani law enforcement. Pakistani officials have at various points conducted guided tours of its school and hospital for the foreign press, presenting the campus as a charitable and educational institution. A network that needed to hide would not have built a two-hundred-acre campus on a main road and invited journalists in. The fuller architecture of the site is examined in the dedicated study of LeT’s headquarters campus, and that openness is the strongest single argument that the network has never needed to conceal its core.

Beyond instruction, Markaz Taiba has functioned as Lashkar’s nerve centre for publishing, propaganda, and missionary outreach. The campus has housed the production of the group’s literature and the administration of its charitable and educational programmes, the soft infrastructure that recruits, raises money, and frames the cause. A facility that is simultaneously a seminary, a hospital, a publishing house, and a training ground cannot be reduced to any single label, and that breadth is part of what made the compound so durable. Striking the drill grounds does not erase the publishing arm, and closing the school does not silence the propaganda. The campus was built as an ecosystem, and ecosystems are harder to kill than camps.

Markaz Subhan Allah at Bahawalpur is the heartland’s second pillar and the property of Jaish-e-Mohammed rather than Lashkar. Spread over roughly fifteen acres on the outskirts of the city, the facility has been assessed as operational since around 2015. It serves as the group’s operational headquarters, its residential quarter for senior figures, and its centre for arms instruction and religious indoctrination. Indian briefers have consistently named it as the site where the Pulwama bombing of February 2019 was planned, and they have placed members of Masood Azhar’s family, including the head of the group’s armed wing, in residence there.

Bahawalpur is not an accidental location for Jaish’s headquarters. The city is Masood Azhar’s hometown, the place he was born and the place his family’s clerical influence runs deepest. When Azhar founded Jaish-e-Mohammed in 2000, after his release in the IC-814 hijacking deal, he built its institutional core where his family already held standing. The Markaz Subhan Allah complex grew alongside a network of seminaries and residences that bound the organisation to the city’s social fabric. This is why the facility could operate so visibly for a decade. It was not an alien installation imposed on a town. It was woven into one, protected by family, by clerical authority, and by the simple fact that thousands of local residents depended on its schools and its charity.

When Indian missiles struck the compound in May 2025, Azhar himself issued a statement claiming that ten members of his family had been killed in the attack. That statement, whatever its propaganda intent, was an unambiguous confirmation that the facility was occupied and that the people occupying it were close to the organisation’s leadership. It also closed, for anyone willing to read it honestly, the long-running argument about whether Markaz Subhan Allah was a genuine militant headquarters or merely a seminary that India had defamed. A seminary does not house the family of a globally designated terrorist leader, and a leader does not mourn a seminary the way Azhar mourned the Bahawalpur strike.

The Sarjal complex at Tehra Kalan is the heartland’s quieter third site, and its design tells a different story from the open campuses of Muridke and Bahawalpur. Located in Shakargarh tehsil of Narowal district, roughly six kilometres from the international boundary, the Sarjal facility functions as Jaish-e-Mohammed’s principal launch point for pushing recruits into Jammu and Kashmir. What distinguishes it is concealment. The site is housed inside the building of a Primary Health Centre in Tehra Kalan village. A government clinic provides cover that a purpose-built compound cannot. Anyone surveilling the structure sees a health facility. Anyone passing through it on the way to the border sees something else.

The choice to embed a launch facility inside a working health centre is not improvisation. It is a calculated exploitation of the protection that civilian institutions enjoy under both domestic law and the laws of armed conflict. A strike on Sarjal can always be presented as a strike on a clinic, and the recruits who pass through it can always be presented as patients or staff. The facility is a small building, but it encodes a large lesson about how the network defends itself: not behind walls, but behind the moral and legal status of the civilian functions it hijacks.

Sarjal is the heartland’s clearest demonstration of the dual-use principle, and it shows that even within Punjab the network can shed its visible footprint when concealment becomes useful.

Taken together, the three Punjab facilities express the network’s institutional confidence. Muridke and Bahawalpur did not hide for twenty-five years and fifteen years respectively because they did not need to. Sarjal hides because a launch point near a contested border benefits from looking like a clinic. The heartland is where the system stores its memory, its leadership, and its reproductive capacity, and it is the layer that the elimination campaign, however many individuals it removed, could never reach by killing one man at a time.

The three Punjab facilities should be read as a single distributed headquarters rather than as three unrelated sites. Muridke holds Lashkar’s institutional memory, Bahawalpur holds Jaish’s, and Sarjal holds the function that neither headquarters wants on its own doorstep, the forward launch role that draws Indian attention. This division of labour means the heartland can absorb a blow without losing everything. A strike on Sarjal removes a launch point but leaves both headquarters intact. A strike on Bahawalpur damages Jaish but leaves Lashkar’s core untouched. The heartland is resilient not because any single site is invulnerable but because the network never concentrated all of its irreplaceable functions in one place.

The heartland’s treatment during the years of financial scrutiny is itself instructive. When Pakistan moved to demonstrate compliance with international demands, the Muridke campus was among the properties it publicly placed under provincial administration, with officials announcing that the schools and the hospital had been brought under government control. The compound did not close. Its boundary wall, its dormitories, and its drill grounds remained, and the announcement amounted to a change of nameplate rather than a change of function. The episode is a preview in miniature of the wider verification dispute examined later in this analysis, and it shows that the heartland facilities have proven able to absorb administrative pressure without surrendering anything structural.

The Camps of Pakistan-Occupied Kashmir

If the Punjab heartland is the network’s memory, the facilities of Pakistan-occupied Jammu and Kashmir are its cutting edge. This zone holds the highest density of sites on the May 2025 map, and the reason is geography. Every facility here exists to move recruits across the Line of Control, and proximity to that line is the variable that organises the entire cluster.

Sawai Nala sits near Muzaffarabad, the territory’s capital, and it belongs to Lashkar-e-Taiba. Also referred to as Bait-ul-Mujahideen, the site has been functional since the early 2000s and handles the front end of the Lashkar pipeline: recruitment, registration, and basic preparation. Indian assessments identify it as one of the facilities where the November 2008 Mumbai attackers were schooled, and they hold that the Pakistan Army provides serving instructors for weapons instruction on its grounds. Sawai Nala is, in effect, the intake gate for a large share of Lashkar’s Kashmir-bound recruits, the point at which a raw arrival becomes a registered cadre.

Markaz Syedna Bilal lies in Muzaffarabad city itself, directly opposite the structure locals call the Red Fort, and it is a Jaish-e-Mohammed property. The facility occupies a three-storey building spread across roughly eight to ten kanals. A cupping-therapy clinic operates on the ground floor. The group’s transit office and a branch of its charitable trust occupy the upper levels, alongside family quarters. The arrangement is the urban version of the Sarjal model. A facility embedded in a city, sharing a wall with a medical clinic and a registered charity, is far harder to characterise, photograph, or strike than an isolated compound on open ground. Syedna Bilal is the network’s demonstration that a forward node does not have to sit in the wilderness.

Around Muzaffarabad lie several further sites that the public map names but that have drawn less attention than Sawai Nala. Chelabandi, Abdullah bin Masood, and Maskar-e-Aqsa belong to this outer Muzaffarabad cluster. Indian assessments treat them as a mix of indoctrination centres and intermediate facilities, sitting between the basic intake of Sawai Nala and the forward launch points closer to the line. Their presence matters for the arithmetic of the network. The Muzaffarabad area is not served by a single camp but by a constellation of them, each handling a different stage or a different group, so that the loss of any one site does not interrupt the flow of recruits through the others.

The Kotli cluster, further south, holds three of the zone’s most significant sites. Gulpur is a Lashkar facility roughly thirty kilometres from the Line of Control, and Indian briefers connected it to militancy across the Rajouri and Poonch belt. The specific attacks attributed to recruits from this facility include the ambush of an army vehicle in the Poonch sector in 2023, in which several soldiers were killed, and the assault on a bus carrying Hindu pilgrims in the Reasi district in June 2024, when gunmen fired on the vehicle and forced it off a mountain road. These were not abstract operations. They were the documented output of a single forward facility, and naming them is what turns Gulpur from a coordinate into a working part of the machine.

Markaz Abbas, around thirteen kilometres from the line and two kilometres from the Kotli military garrison, is a Jaish facility assessed as able to house between a hundred and a hundred and twenty-five recruits, with forty to fifty present at any given time. It is run by a founding member of Jaish who sits on the group’s governing council, a man who was himself part of the older Harkat-ul-Mujahideen before the founders broke away to form Jaish around Masood Azhar. The compound’s proximity to a regular army camp is the single most damaging fact in the Kotli file, because it places a recruit-housing facility for a banned organisation within a short walk of a serving military garrison.

Maskar Raheel Shaheed, the third Kotli site, belongs to Hizbul Mujahideen, ranks among the oldest facilities in the zone, and is assessed as capable of sheltering a hundred and fifty to two hundred recruits with specialised instruction in sniping, ambush, and mountain warfare. Where Sawai Nala is an intake gate and Markaz Abbas is a mid-sized residential compound, Raheel Shaheed is closer to a finishing school, a place where recruits who have already absorbed the basics are given the marksmanship and fieldcraft suited to a sustained campaign in the forests and ridgelines of Indian Kashmir.

Markaz Ahle Hadith Barnala completes the cluster. Situated near Bhimber, on the outskirts of Barnala town along the Kote Jamel road, this Lashkar facility sits roughly nine kilometres from the Line of Control, the closest of any site on the public map. Its function is infiltration into the Poonch, Rajouri, and Reasi sectors, moving both recruits and weapons. Its position underlines the organising logic of the whole zone. The deeper a facility sits, the more it resembles a school. The closer it sits to the line, the more it resembles a departure lounge. Barnala, nine kilometres out, is almost entirely a departure lounge.

The Kashmir zone also reveals how the forward facilities are designed to be replaceable. Unlike the two-hundred-acre campus at Muridke, a forward node is typically a modest building, a rented compound, or a few rooms above a shop. It can be stood up in weeks, abandoned under pressure, and re-established a few kilometres away with comparatively little loss. The detachments and launch pads that surround the named markaz sites are even more transient. This is why the count of facilities in Pakistan-occupied Kashmir is inherently unstable. A satellite chart captures the permanent compounds. It cannot keep pace with the smaller forward sites that shift with the infiltration season and with the pressure of the moment.

Seasonality governs the forward zone as firmly as geography does. The high passes of the Line of Control are most readily crossed in the warmer months, when snow has cleared from the routes and foliage gives cover to small groups moving at night. Indian assessments describe an infiltration calendar that builds through spring, peaks in summer and early autumn, and slackens as winter seals the passes. The forward facilities breathe with this calendar. A site near Bhimber that holds a thin population through the cold months fills as the season opens, and the launch pads closest to the line are activated only when a crossing is imminent. A satellite chart taken in February and a chart taken in July can therefore describe what looks like two different networks, when in fact they describe one network at two points in its annual rhythm.

Mosques recur throughout the Kashmir zone not as incidental neighbours of the camps but as working parts of them. Markaz Syedna Bilal is built around a mosque. Several of the forward sites use a mosque as a gathering point, a place where recruits can assemble without drawing notice and where a transit office can operate on an upper floor. The religious building provides three things the network values: a legitimate reason for groups of young men to congregate, a measure of legal and moral protection against scrutiny, and a setting in which indoctrination feels like worship. The mosque is, in the forward zone, what the health centre is at Sarjal, a civilian institution whose protected status the network has learned to borrow.

The Kashmir zone is where the network meets the border, and the routes its graduates use to cross are mapped in detail in the study of Line of Control infiltration corridors. What the zone reveals about the network as a whole is its modularity. A heartland campus can take years to build and decades to mature. A forward node in a rented building opposite a city landmark can be stood up, abandoned, and re-established with comparatively little effort. The forward zone is designed to be expendable in a way the heartland is not, and that expendability is why striking it, on its own, would never have satisfied Indian planners.

The Khyber Pakhtunkhwa Cluster

The third zone of the network runs through the Mansehra district of Khyber Pakhtunkhwa, and it is the layer with the longest history. Garhi Habibullah, Batrasi, Oghi, Boi, and the camp at Balakot all fall within this belt, and the belt’s character is shaped by terrain, by history, and by a degree of remoteness that the open Punjab plains do not offer.

Mansehra became an instruction corridor during the 1980s. When the United States, Saudi Arabia, and Pakistan ran the covert war against the Soviet occupation of Afghanistan, the districts along the Afghan frontier filled with training grounds, weapons depots, and transit routes for fighters drawn from across the Muslim world. That infrastructure did not disappear when the Soviet army withdrew in 1989. It was repurposed. The same camps, the same instructors, and the same logistics chains were turned from the Afghan jihad toward Kashmir during the early 1990s. The Mansehra belt is, in this sense, the oldest continuously operating segment of the network, and its lineage runs straight back to a war that ended before most of its current recruits were born. The deeper history of how the Afghan conflict seeded Pakistan’s armed groups belongs to a separate discussion, but the camps of Mansehra are its most physical inheritance.

The belt’s character cannot be separated from the population that surrounds it. Mansehra and the adjoining districts absorbed large numbers of Afghan refugees during and after the Soviet war, and the refugee economy created exactly the conditions a recruitment pipeline thrives on: displaced families, weak documentation, religious schools offering food and shelter where the state offered little, and a steady supply of young men with few prospects. The instruction belt did not have to be imposed on an unwilling region. It grew within a society that the long Afghan emergency had already reshaped, and that organic embedding is part of why it has proven so durable.

Balakot is the belt’s most internationally recognised name, and for a specific reason. The town and the Jaba ridge above it hosted a Jaish-e-Mohammed facility that became the target of India’s February 2019 airstrike, the operation launched in response to the Pulwama bombing. In the early hours of February 26, 2019, Indian Mirage jets crossed into Pakistani airspace and released precision munitions at the hillside compound. That strike marked the first time since the 1971 war that Indian aircraft had crossed into Pakistani territory proper to bomb a target, and the camp at Balakot was the reason.

The Balakot strike also previewed every argument that would later attend Operation Sindoor. India stated that the operation had destroyed a functioning facility and killed a large number of recruits. Pakistan stated that the bombs had fallen on a forested hillside and harmed nothing but trees. Independent verification was thin, and the dispute hardened into the familiar pattern in which each government’s account served its domestic audience. What was not in dispute, and what mattered more than the body count, was the doctrinal precedent. India had treated a training facility as a legitimate object for a cross-border military strike. The 2019 operation is therefore the network’s first documented case of a compound, rather than an individual, becoming the object of an air strike, and it was a preview, six years early, of the doctrine that Operation Sindoor would apply at far greater scale.

Garhi Habibullah, Batrasi, Oghi, and Boi, the other Mansehra sites, are less famous but structurally important. They provide what intelligence assessments describe as mid-tier and specialised preparation: courses beyond the basic intake handled at forward nodes such as Sawai Nala, set in valleys and ridgelines that offer concealment from routine observation. Garhi Habibullah, near the boundary between Khyber Pakhtunkhwa and Pakistan-occupied Kashmir, has long been associated with the movement of recruits between the two zones, while Batrasi and the higher camps offer the terrain for endurance and mountain conditioning. The belt functions as the network’s middle school, positioned between the basic instruction of the Kashmir zone and the institutional core of the Punjab heartland. Its terrain is its protection. A compound on the Punjab plain is visible from orbit at a glance. A facility tucked into a forested Mansehra valley demands far more deliberate surveillance to confirm.

The Khyber Pakhtunkhwa belt also connects the anti-India network to a wider and more chaotic militant landscape. The province has hosted not only Kashmir-focused groups but the Afghan Taliban’s logistics, the Pakistani Taliban’s insurgency, and a tangle of sectarian outfits. The geography that shelters one set of camps shelters the others, and the tribal-belt terror landscape and the lawless zones of Waziristan form the broader environment in which the Mansehra facilities sit. The belt is not a sealed compartment. It is the seam where the India-directed network touches every other armed project Pakistan’s territory has hosted.

The Balakot case taught both sides a lesson that shaped what came after. India learned that a single strike on a single facility, however precise, could be absorbed and denied, its effect blunted by the absence of independent verification and by Pakistan’s control of the site afterward. Pakistan learned that the heartland was not as far from Indian reach as it had assumed, even though the 2019 strike sat at the edge of the network rather than its core. Both lessons fed directly into 2025. India planned Operation Sindoor as a multi-site campaign precisely because one site had proven deniable, and it presented its evidence in advance precisely because the Balakot aftermath had been a contest of unverified claims. The Mansehra belt’s most famous compound thus shaped the doctrine that would later strike the whole network.

Read as a whole, the Khyber Pakhtunkhwa belt performs a strategic function that neither the heartland nor the forward zone can replicate. It is the network’s reserve of depth and concealment. If pressure on Punjab ever forced the heartland to disperse, the forested valleys of Mansehra offer the terrain to absorb that dispersal. If the forward zone in Kashmir is degraded, the belt offers a sheltered area to rebuild and re-train before pushing recruits forward again. Its value lies precisely in being unglamorous and hard to see, the part of the system that does not feature in dramatic briefings but that gives the whole network somewhere to retreat to. A campaign that ignored the belt would leave the network its safest room.

Verifying activity in the Mansehra belt is harder than anywhere else in the network. Where a Punjab campus can be photographed from orbit in a single clear frame, a facility folded into a forested valley demands repeated, deliberate observation across seasons to confirm what it is. Cloud cover, tree canopy, and the ordinary rhythms of mountain village life all blur the picture. This is why the belt’s site count has always been the softest part of any public map, and why assessments of the zone lean more heavily on human sources than on imagery. The terrain that makes the belt valuable to the network is the same terrain that makes it opaque to the analysts trying to watch it.

When the Camps Became Targets

For most of the network’s history, the facilities themselves were never struck. India’s response to terrorism originating in these compounds took other forms: diplomatic protest, dossiers submitted to the United Nations, pressure through the Financial Action Task Force, and, from 2021 onward, the quiet elimination of camp graduates inside Pakistani cities. The buildings stood untouched. That changed in stages, and the change is the most consequential development in the network’s recent history.

The first crack came in February 2019. After a Jaish-e-Mohammed suicide bomber killed forty Indian paramilitary personnel at Pulwama, Indian aircraft crossed the international boundary and bombed the Jaish facility on the Jaba ridge above Balakot. The damage assessment from that strike was disputed for years, with India claiming the destruction of infrastructure and Pakistan claiming the bombs fell on an empty hillside. The military significance, however, did not depend on the body count. For the first time, India had treated a training facility as a legitimate target for a cross-border air operation. The principle had been established. Only the scale remained to be decided.

The scale was decided on the night of May 7, 2025. The trigger was the massacre at Pahalgam on April 22, when gunmen opened fire on tourists at the Baisaran meadow in Kashmir, killing twenty-five Indians and one Nepali citizen. India attributed the attack to The Resistance Front, a label its assessments treat as a front for Lashkar-e-Taiba. The fortnight between the massacre and the strike filled with escalating Indian measures, from the suspension of the Indus Waters Treaty to a wave of diplomatic steps, and it ended with a decision to strike the network directly.

In the early hours of May 7, Indian Rafale jets, supported by the navy’s precision munitions, struck nine sites across Pakistan and Pakistan-occupied Kashmir. The operation was codenamed Sindoor, and the granular reconstruction of those minutes is laid out in the account of the Sindoor strikes on training facilities. The aircraft fired stand-off weapons from within Indian airspace, a method that allowed the strike to reach deep targets without the jets themselves crossing the border, and the operation was compressed into a window of roughly half an hour.

The target list was the network itself. Five of the nine sites lay in Pakistan-occupied Kashmir: Sawai Nala and Markaz Syedna Bilal at Muzaffarabad, Gulpur and Markaz Abbas in the Kotli area, and Markaz Ahle Hadith Barnala near Bhimber. The remaining four lay inside Pakistan proper: Mehmoona Joya near Sialkot, the Sarjal complex at Tehra Kalan in Shakargarh, Markaz Taiba at Muridke, and Markaz Subhan Allah at Bahawalpur. Every one of the nine appears on the inventory described in the preceding sections. The Indian Director General of Military Operations, Lieutenant General Rajiv Ghai, and the Director General of Air Operations, Air Marshal A. K. Bharti, told reporters that the target systems handed to the air force were, in Bharti’s words, the notorious compounds at Bahawalpur and Muridke, and that critical target-system analysis had been used to plan their destruction.

The selection of those nine sites from the wider catalogue was itself a deliberate signal. India did not strike every facility it had mapped. It struck a representative spread, reaching into all of the network’s zones at once: the forward Kashmir cluster, the Sialkot border facility, and the two heartland campuses that anchor the whole structure. Indian briefers reinforced the message by presenting pre-strike imagery of the targeted compounds, walking reporters through the buildings as they had stood before the missiles arrived. The point of that presentation was to pre-empt the denial that had followed Balakot. By placing the evidence on the record before Pakistan could claim the strikes had hit empty ground, India was arguing not only that the camps existed but that their existence had been documented well in advance.

Two features of the May operation define its meaning for the network. The first is that India struck the heartland, not merely the forward edge. Hitting Barnala or Gulpur would have damaged the expendable layer. Hitting Muridke and Bahawalpur reached the institutional core, the campuses that the network had operated openly for decades on the assumption that they were beyond military reach. That assumption died on May 7. The second feature is the self-imposed restriction the Indian command described. Bharti and Ghai stated that the strikes were confined to terror infrastructure and deliberately avoided Pakistani military facilities, a choice intended to keep the operation focused and to limit the escalation. Whether that restraint held through the days that followed is a separate question, because Pakistan’s response and the subsequent exchanges expanded the conflict well beyond the original nine sites.

In the hours and days after the strike, Pakistan launched waves of drones and missiles toward Indian military installations, and India answered by striking Pakistani radar sites and air bases across the western front. The crisis climbed several rungs of the escalation ladder before a ceasefire was arranged on May 10. The point for the present analysis is narrower. The opening move of that whole crisis, the act that began it, was a strike on the training network. The camps were not a side target. They were the reason the crisis happened, the object whose continued existence had made the wider confrontation feel inevitable to Indian planners.

The strike on Mehmoona Joya, the Hizbul Mujahideen facility near Sialkot, deserves note because it widened the operation beyond the two dominant groups. Lashkar and Jaish had drawn most of the attention after Pahalgam, and most of the nine targets belonged to them. Including a Hizbul site signalled that the operation was aimed at the network as a category rather than at two organisations as defendants. Indian briefers connected Mehmoona Joya to the planning of the 2016 Pathankot air base assault, reaching back nearly a decade to anchor the target in a documented attack. The choice underlined the operation’s logic. Sindoor was not retaliation against the single group behind one massacre. It was an attack on the system that produces such groups, and a system has to be hit across all of its parts.

Damage assessment remains contested, as it was after Balakot. India stated that more than a hundred militants were killed across the nine facilities. Pakistan said the strikes killed thirty-one civilians and hit mosques and inhabited areas. Independent verification is limited, and commercial satellite imagery, which can show structural damage and reconstruction, cannot count the dead or confirm the affiliation of those present. What the imagery and the public record can establish is narrower and still significant. The buildings existed. They were occupied. Masood Azhar’s own statement mourning family members killed at Bahawalpur removed any doubt that the Subhan Allah facility was a live, populated site rather than the empty shell Pakistan’s broader narrative had long implied. The strike that destroyed the camps also, inadvertently, proved the camps were real.

Body counts after a strike of this kind are notoriously unreliable, and the dispute over Sindoor’s toll is, in a sense, structurally unresolvable. Satellite imagery records buildings, not bodies. Each government has a powerful incentive to shape its figure: India to demonstrate that the operation succeeded, Pakistan to present it as an atrocity against civilians. No neutral investigator had access to the sites in the aftermath. What can be said with confidence is bounded and still meaningful. The facilities were real, they were occupied, and they were struck. Beyond that, the precise human toll belongs to the category of contested claims, and an honest account marks it as such rather than adopting either side’s arithmetic.

The deeper meaning of Operation Sindoor lies in the doctrinal line it crossed. For decades, the Indian state had treated the camps as a problem to be managed through pressure on Pakistan rather than as targets to be destroyed directly. The shadow war of targeted eliminations, which began in 2021, had already shifted that posture by reaching into Pakistani cities to kill the network’s graduates. The May operation completed the shift. It announced that the facilities themselves, including the heartland campuses, were now inside the target set, and it told every compound still standing that the old immunity was gone.

Operational or Shut Down: The Verification Problem

The central dispute about Pakistan’s instruction network is not where the facilities sit. Their locations have been mapped, photographed, and published. The dispute is whether they are still working. Pakistan’s position and the assessment of Indian and Western intelligence diverge sharply on this point, and the divergence deserves to be examined honestly rather than resolved by assertion.

Pakistan’s claim rests on the Financial Action Task Force years. Between 2018 and 2022, the international financial watchdog placed Pakistan on its grey list, the enhanced-monitoring category reserved for states judged deficient in countering terror financing and money laundering. Grey-listing is not a sanction in the ordinary sense. What it does is raise the cost and friction of a country’s access to international finance, signalling to banks and investors that the state carries elevated risk. For a Pakistani economy already under strain, that signal carried real weight, and the grey-list years coincided with a genuine and visible scramble in Islamabad to be seen taking action.

During that period, Pakistan took a series of demonstrable steps. It seized properties linked to Jamaat-ud-Dawa, the Lashkar front. It prosecuted Hafiz Saeed on terror-financing charges and secured convictions carrying multi-year sentences. It announced the takeover of seminaries, dispensaries, and ambulances tied to the proscribed groups. It updated its schedule of banned organisations and publicised arrests. In October 2022, the task force removed Pakistan from the grey list, a decision Islamabad presented as international validation that its camps had been dismantled and its militant infrastructure brought under control.

The Indian and Western assessment reads the same period very differently. In this account, the FATF years produced rebranding rather than dismantlement. The properties seized were transferred, on paper, to provincial authorities while the same personnel continued to use them. The convictions of senior figures coincided with continued comfortable detention rather than genuine incapacitation. The proscribed groups acquired new names, The Resistance Front and the People’s Anti-Fascist Front prominent among them, designed specifically to let attacks proceed without triggering the scrutiny attached to the old labels. The camps, in this reading, did not close. They lowered their profile, dispersed some functions, and in several cases moved deeper into civilian cover, the Sarjal facility inside a Primary Health Centre being the clearest example of a site built to survive exactly this kind of pressure.

The relabelling of attacks under fresh group names served the same purpose at the level of operations that the embedding of facilities inside clinics served at the level of real estate. Both tactics were designed to let the network keep functioning while presenting a compliant surface to international monitors. The grey-list exit of October 2022, in this view, certified a performance rather than a transformation, rewarding Pakistan for the appearance of action without confirming that the instruction system had actually been dismantled.

There is also a structural reason the verification problem is so stubborn. The network’s most resilient facilities are precisely the ones built to defeat verification. A purpose-built compound on open ground is easy to assess and easy to strike. A launch pad embedded in a health centre, a transit camp sharing a wall with a charity clinic, a detachment of rented rooms in a town: these are designed so that no single observation can prove what they are. The harder the network is squeezed, the more it migrates toward this ambiguous category, which means the very pressure meant to shut it down also makes it harder to confirm whether it has been shut down. Verification and concealment are locked in a contest, and the network has spent two decades learning to win it.

Adjudicating between these accounts requires being precise about what evidence can and cannot show. Commercial satellite imagery is powerful but limited. It can reveal whether a compound still stands, whether new construction has occurred, whether vehicles cluster at a site, and whether a facility destroyed in a strike has been rebuilt. It cannot, on its own, confirm that a given building ran a weapons course last month, because a dormitory, a drill ground, and a religious seminary look broadly similar from orbit. Satellite analysis narrows the question. It does not close it.

The dispute has always had an international dimension as well as a bilateral one. India has for years submitted dossiers to the United Nations and lobbied the sanctions committee that maintains the global list of designated terrorists, securing the listing of figures such as Masood Azhar after prolonged diplomatic effort. Western governments and the financial task force have at various points endorsed parts of the Indian case, which is what produced the grey-listing in the first place. Yet international designation has consistently outrun international enforcement. A man can be named on a sanctions list and still run a seminary, and a country can be grey-listed and still host the seminary. The gap between what the international system has formally recognised and what it has actually changed on the ground is, in the end, the same gap that Operation Sindoor was an attempt to close by force.

Decisive evidence, in the end, came not from imagery but from Operation Sindoor itself. If Pakistan’s account were accurate, if the camps had genuinely been shut down and emptied during the FATF years, then the missiles of May 7 struck nine abandoned buildings. That is not what the record shows. India presented pre-strike imagery of active facilities. Masood Azhar publicly mourned family members killed at Bahawalpur, an admission that the Subhan Allah headquarters was populated by people close to the Jaish leadership. Pakistan’s own casualty claims, even in their civilian-focused framing, conceded that the struck sites were occupied. A network of empty buildings does not produce casualties when it is bombed. The May 2025 strike functioned, among other things, as the most authoritative verification exercise the dispute has ever received, and its finding was unambiguous. The facilities were live.

The honest conclusion is therefore a qualified one. Pakistan’s claim that nothing changed during the FATF years is false. Things did change. Visibility dropped, some properties shifted nomenclature, and the network adapted to financial scrutiny with genuine effort to look compliant. But the claim that the camps were dismantled is equally false, because a dismantled network cannot be struck and cannot bleed. What the evidence supports is adaptation rather than abolition. The instruction system the wider shadow war’s broader infrastructure depends on did not vanish under pressure. It changed shape and kept working, which is the behaviour of an institution being protected, not an institution being dismantled.

This conclusion also explains why the financial process, for all its real economic bite, could never resolve the problem on its own. A financial watchdog can audit money flows, demand legislation, and check boxes. It cannot peer inside a Primary Health Centre in Tehra Kalan and determine who is a patient and who is a recruit. The grey-list mechanism measured Pakistan’s compliance with a procedural standard, and Pakistan learned to meet that standard without surrendering the network. The gap between procedural compliance and actual dismantlement is the space in which the camps survived, and Operation Sindoor was, in part, an admission that no checklist would ever close that gap.

The months after Operation Sindoor offered their own data point. Commercial satellite imagery of several struck sites showed clearance of rubble and signs of reconstruction work, and Pakistani authorities moved quickly to control access to the damaged compounds, restricting journalists and managing what could be photographed. Reconstruction is itself a form of evidence. A genuinely abandoned facility is left as ruins, because there is nothing to rebuild it for. A facility that matters is cleared and rebuilt. The early signs that the network was repairing rather than abandoning its struck sites suggested that the institutions valued the addresses enough to invest in restoring them, which is not the behaviour of a system that had already been dismantled.

What the Network Produces and How the Shadow War Changed It

A map of buildings becomes meaningful only when it is read as a measure of output. The relevant question about the instruction network is not how many compounds exist but how many fighters they generate, because that figure is what the wider campaign against Pakistan-based terrorism is ultimately trying to reduce.

The clearest single anchor for an estimate is Markaz Taiba. Intelligence assessments place the annual throughput of the Muridke campus at roughly a thousand recruits across its various courses. If the heartland’s largest single facility processes a thousand a year, the full network, with its forward nodes in the Kashmir zone and its specialised sites in the Mansehra belt, plausibly graduates several thousand annually, though the figure cannot be stated with precision and varies with the intensity of each year’s recruitment drive. The number that matters is comparative rather than absolute. The shadow war’s elimination campaign has removed a documented total of wanted figures that runs into the dozens since 2021. The network’s annual graduating capacity runs into the thousands. The arithmetic gap between those two figures is the single most important fact in the entire analysis, because it explains why Indian planners stopped treating individual eliminations as a complete strategy.

Instruction itself is structured, not improvised. Lashkar-e-Taiba’s course system, the best-documented of the three groups, runs in recognised stages. A preliminary religious phase, conducted at seminary level, screens and indoctrinates new arrivals. A basic course of roughly three weeks, sometimes described by the group’s own term Daura-e-Aam, delivers small-arms familiarity and physical conditioning. An advanced course of around three months, the Daura-e-Khaas, adds explosives handling, ambush tactics, survival skills, and reconnaissance methodology. Surveillance-specific instruction prepares recruits to study targets before an operation. The men who attacked Mumbai in 2008 moved through this sequence. The structure means the network is not a collection of rifle ranges. It is a graded curriculum, and the seminary screening that feeds its first stage connects directly to the recruitment feeder system that converts impoverished students into committed cadres long before they ever reach a drill ground. A facility without a feeder pipeline would run out of recruits, and the madrassa network is that pipeline, the mechanism that delivers raw human material to the gate of the first course.

The throughput also explains why the network functions as a reproduction mechanism rather than a fixed arsenal. An armoury can be emptied. A trained cohort, once deployed and lost, would leave a permanent hole if there were no way to replace it. The facilities exist precisely to make the loss of a cohort temporary. Each season’s intake replaces the previous season’s casualties and graduates, and the system is designed so that attrition, however severe, is met by reproduction. This is the deepest reason the elimination campaign alone could never win. Killing graduates is attrition, and attrition against a system built for reproduction is a race the camps were engineered to win.

Cost is part of why reproduction works so well for the network. Producing a trained fighter is cheap compared with the disruption that fighter can cause. A recruit drawn from a poor district, fed and housed through a seminary, and moved through a few months of instruction represents a modest outlay for the organisation, while a single successful attack can shape headlines, policy, and the security posture of an entire region. This asymmetry, low cost of production against high impact of output, is the grim economic logic that has kept the camps running. It also explains why financial pressure alone, however well aimed, struggles to bite. The network does not need to be rich. It needs only to be cheaper to run than the damage it inflicts.

The shadow war has changed this network, though not in the way a casual reading might assume. The campaign of targeted eliminations that began in 2021 did not shut a single camp. What it did was raise the cost of being a known figure inside Pakistan. Camp commanders, instructors, and senior planners who once moved openly began to alter their routines, vary their residences, and accept the security restrictions that come with being hunted. Several known figures connected to the network have died in unexplained shootings, in incidents Pakistani authorities have variously attributed to unidentified gunmen or to internal disputes. The effect on the facilities was indirect but real. A network whose senior personnel must now behave like wanted men operates with more friction, more caution, and more attention diverted to self-protection than a network whose leadership feels untouchable. The hunt did not break the system. It made the system nervous, and a nervous network is a less efficient one.

Operation Sindoor changed the network far more directly. For the first time, the campuses themselves, including the heartland anchors at Muridke and Bahawalpur, were physically struck. The immediate effect was destruction and disruption. The deeper effect was the removal of an assumption. For twenty-five years the network had operated on the premise that its core facilities, sitting deep inside Punjab, were beyond the reach of anything but diplomatic protest. That premise is gone. Every compound on the map now operates in the knowledge that it can be reached, and that knowledge imposes its own tax on the system: more dispersal, smaller footprints, greater investment in concealment and in moving functions into civilian cover.

What permanent degradation would actually require is worth stating plainly, because it clarifies the limits of any military option. The facilities are physical, and physical things can be destroyed and, given protection and money, rebuilt. The instructors are people, and people can be killed and, given a recruitment base, replaced. The one input that is neither a building nor a person is the political decision in Rawalpindi to keep the system alive. As long as that decision holds, every strike buys time rather than victory, because the state that built the network will fund its reconstruction. Permanent degradation is therefore less a military problem than a political one, and no quantity of missiles can substitute for a change of mind in the institution that has spent three decades treating these camps as instruments of policy.

The network’s output does not stay within the India-Pakistan frame. Skills taught for a Kashmir campaign, the handling of explosives, the planning of urban assaults, the methods of reconnaissance and concealment, are transferable, and graduates of these facilities have surfaced in plots and attacks well beyond Kashmir over the decades. The instruction system is a regional security problem rather than a purely bilateral one, which is why Western intelligence services have tracked it alongside their Indian counterparts and why the financial task force, a global body, was drawn into the matter at all. A factory that manufactures fighters cannot guarantee where its products are eventually used, and that is one reason the camps have drawn attention far from the Line of Control.

Taken together, the shadow war and Operation Sindoor represent two halves of a single logic rather than two separate strategies. The eliminations addressed the network’s output, the trained individuals, by removing them after they had graduated. The strikes addressed the network’s capacity, the facilities, by attacking the places where graduates are made. Neither half is sufficient alone. Eliminations without strikes leave the factory running. Strikes without eliminations leave the existing graduates at large. The escalation from one to the other was not a replacement but an addition, an acknowledgement that a problem with both a stock and a flow has to be attacked on both fronts at once.

Yet the network’s three-zone depth, the very feature described at the start of this analysis, is also the reason a single operation cannot end it. Sindoor struck nine sites. The May 2025 map showed twenty-one, and intelligence assessments have over the years identified others not on the public chart, along with the launch pads and detachments that no chart can fix permanently. A heartland campus can be rebuilt. A forward node in a rented building can be re-established within a season. The instruction system is wounded, exposed, and demonstrably no longer immune, but it is not destroyed, and an honest account must hold both of those facts at once.

This is the structural truth the camp map finally makes visible. The shadow war’s eliminations subtract individuals, one graduate at a time, from a system designed to replace them by the thousand. Operation Sindoor attacked the replacement mechanism itself, the factories rather than the products. That escalation was not a change of mood. It was a recognition, forced by the arithmetic, that a campaign confined to hunting people could never close the gap between how fast Pakistan’s network produces fighters and how fast its enemies can remove them. Whether the network can be permanently degraded depends on a question no strike can answer, the question of whether the state that built it ever decides to stop protecting it. Until that decision is made, the map of training facilities is a map of why the conflict had to escalate, and a map of how much of the network is still standing after it did.

Frequently Asked Questions

How many terror training camps exist in Pakistan?

There is no single agreed figure, because the count depends on what is classified as a camp and on which year’s assessment is used. The most authoritative public number is the twenty-one facilities India displayed on a map at its Operation Sindoor briefing on May 7, 2025, spread across Pakistan and Pakistan-occupied Kashmir. Indian and Western intelligence assessments have over the years identified additional sites not on that public chart, including launch pads and detachments too small or too transient to map permanently. The honest answer is that the publicly verified core of the network numbers around twenty-one major facilities, while the full system, counting forward nodes and temporary sites, is larger and harder to fix at a precise number.

Where are Pakistan’s known training camps located?

The facilities cluster into three zones. The Punjab heartland holds Markaz Taiba at Muridke near Lahore, Markaz Subhan Allah at Bahawalpur, and the Sarjal complex at Tehra Kalan in Shakargarh. Pakistan-occupied Jammu and Kashmir holds the densest cluster, including Sawai Nala and Markaz Syedna Bilal at Muzaffarabad, Gulpur and Markaz Abbas near Kotli, and Markaz Ahle Hadith Barnala near Bhimber. The Mansehra belt of Khyber Pakhtunkhwa holds Garhi Habibullah, Batrasi, Oghi, Boi, and the Balakot camp. The three zones perform different functions, with the heartland serving as institutional core and the Kashmir cluster serving as the forward edge nearest the Line of Control.

What training do the camps provide?

Instruction is structured in graded stages rather than delivered as a single block. The best-documented system, run by Lashkar-e-Taiba, begins with a religious and ideological phase at seminary level, moves to a basic course of roughly three weeks covering small-arms handling and physical conditioning, and culminates in an advanced course of around three months that adds explosives handling, ambush tactics, survival skills, and reconnaissance methodology. Specialised facilities such as Maskar Raheel Shaheed near Kotli add sniping and mountain warfare. Forward nodes near the Line of Control concentrate on the final stages of preparation and on the logistics of crossing the border. Not every recruit completes every stage. The seminary phase screens out many, and only a portion are selected for the advanced course or for specialised tracks such as sniping or reconnaissance. The system is therefore a filter as much as a school, narrowing a broad intake down to the smaller number assessed as suited to operations.

Were training camps targeted during Operation Sindoor?

Yes. Operation Sindoor, conducted on the night of May 7, 2025, struck nine sites, all of them part of the instruction network. Five lay in Pakistan-occupied Kashmir: Sawai Nala and Markaz Syedna Bilal at Muzaffarabad, Gulpur and Markaz Abbas near Kotli, and Markaz Ahle Hadith Barnala near Bhimber. Four lay inside Pakistan proper: Mehmoona Joya near Sialkot, the Sarjal complex in Shakargarh, Markaz Taiba at Muridke, and Markaz Subhan Allah at Bahawalpur. The Indian military stated the targets were terror infrastructure and that Pakistani military facilities were deliberately avoided.

Has Pakistan shut down any training camps?

Pakistan claims it dismantled its militant infrastructure during the years it spent on the Financial Action Task Force grey list, between 2018 and 2022, pointing to property seizures, the prosecution of Hafiz Saeed, and the announced takeover of seminaries. Indian and Western assessments disagree, arguing that the period produced rebranding and concealment rather than genuine closure. The strongest single piece of contrary evidence is Operation Sindoor itself, because the May 2025 strikes hit occupied facilities and produced casualties, including family members of the Jaish leadership at Bahawalpur. A genuinely dismantled network cannot be struck and cannot bleed, which indicates adaptation rather than abolition. Pakistan’s own conduct undercuts its claim in another way. It has continued to prosecute and detain figures it simultaneously insists run only charities, an inconsistency that is difficult to reconcile with the position that the militant infrastructure no longer exists.

Can satellite imagery confirm which camps are active?

Only partially. Commercial satellite imagery is valuable for confirming that a compound still stands, detecting new construction, observing vehicle activity, and verifying whether a facility destroyed in a strike has been rebuilt. What it cannot do is confirm that a particular building ran a weapons course in a particular month, because a dormitory, a drill ground, and a religious seminary appear broadly similar from orbit. Imagery narrows the question of which sites are active but does not settle it on its own. Confirmation generally requires human intelligence and intercepted communications alongside the overhead view. Imagery is most decisive in the negative direction. It can show clearly when a struck facility has been cleared and rebuilt, which is itself evidence that the site still matters to its operators. What it cannot do is read intent, and intent is what separates a functioning camp from an idle compound.

Which organisation runs the most training camps?

Lashkar-e-Taiba operates the largest share of the network. Its facilities span all three geographic zones, from the institutional anchor at Markaz Taiba in Muridke to forward nodes such as Sawai Nala, Gulpur, and Barnala. Jaish-e-Mohammed runs the second cluster, anchored by Markaz Subhan Allah in Bahawalpur and including Markaz Syedna Bilal, Markaz Abbas, and the Sarjal launch complex. Hizbul Mujahideen holds the third position, with Mehmoona Joya near Sialkot and Maskar Raheel Shaheed near Kotli. Behind all three groups stands the Pakistani intelligence directorate, which Indian and Western assessments hold supplies instructors and protection. Lashkar’s dominance is partly a function of its size and partly of its reach. Because its facilities span all three zones, from the heartland to the forward edge, the group is less dependent on any single site than its rivals, and that distribution is one reason it has remained the network’s principal operator for three decades.

How many fighters can the camp network produce annually?

A precise figure is impossible to state, but the network’s largest single facility provides an anchor. Markaz Taiba at Muridke is assessed to process roughly a thousand recruits a year across its courses. Adding the forward nodes in the Kashmir zone and the specialised sites in the Mansehra belt, the full network plausibly graduates several thousand recruits annually, with the exact total varying by year and recruitment intensity. The figure that matters is comparative. The network’s annual output runs into the thousands, while the shadow war’s elimination campaign has removed wanted figures numbering in the dozens, a gap that explains why India escalated to striking the facilities themselves. The figure should be read as a capacity rather than a fixed output. Recruitment rises and falls with political conditions, with the intensity of the Kashmir campaign, and with the pressure the network is under. The camps can run below capacity for years and then surge, which is what makes the instruction system so difficult to neutralise through attrition alone.

What is Markaz Taiba in Muridke?

Markaz Taiba is the headquarters campus of Lashkar-e-Taiba, established in 2000 at Nangal Sahdan in the Muridke area of Sheikhupura district, on the outskirts of Lahore. The complex spreads across a campus estimated at roughly two hundred acres and contains a mosque, a hospital, a school, dormitories, administrative offices, and, by intelligence account, weapons storage and tactical drill areas. Around a thousand recruits are assessed to pass through its courses each year. The site is best known as the facility where the perpetrators of the November 2008 Mumbai attacks were prepared, and it operated openly for twenty-five years before being struck during Operation Sindoor.

What is Markaz Subhan Allah in Bahawalpur?

Markaz Subhan Allah is the operational headquarters of Jaish-e-Mohammed, located on roughly fifteen acres on the outskirts of Bahawalpur in Pakistani Punjab and assessed as operational since around 2015. The facility serves as the group’s centre for arms instruction and religious indoctrination and as a residence for senior figures, including members of Masood Azhar’s family. Indian briefers have consistently linked it to the planning of the February 2019 Pulwama bombing. When Indian missiles struck the compound in May 2025, Azhar issued a statement claiming ten members of his family had been killed there, confirming that the site was occupied.

Was Ajmal Kasab trained at a Pakistani camp?

Yes. Ajmal Kasab, the sole attacker captured alive after the November 2008 Mumbai assault, received his preparation within the Lashkar-e-Taiba network. Indian briefings have identified Markaz Taiba at Muridke and the Sawai Nala facility near Muzaffarabad as sites connected to the schooling of the Mumbai attackers. The planners of that operation, including figures later named in international investigations, reportedly visited the Muridke campus. Kasab’s path through the network is one of the most thoroughly documented illustrations of how the instruction system converts a recruit into an operational attacker.

What is the difference between a training camp and a launch pad?

The two perform different roles in the same pipeline. A training camp delivers instruction, taking recruits through religious indoctrination, basic preparation, and advanced courses, and the heartland campuses at Muridke and Bahawalpur are examples. A launch pad sits close to the Line of Control and exists to stage recruits for the final movement across the border, providing last briefings, equipment, and a departure point rather than a full curriculum. Forward nodes such as Markaz Ahle Hadith Barnala, roughly nine kilometres from the line, function largely as launch pads. Many facilities blend both roles, but the deeper a site sits, the more it resembles a school, and the closer it sits to the line, the more it resembles a departure point.

Did the 2019 Balakot strike hit a training camp?

Yes. India’s airstrike of February 2019, launched in response to the Pulwama suicide bombing, targeted a Jaish-e-Mohammed facility on the Jaba ridge above Balakot in the Mansehra district of Khyber Pakhtunkhwa. The damage assessment was disputed, with India claiming destruction of infrastructure and Pakistan claiming the bombs fell on empty ground. The military significance, however, was that India for the first time treated a training facility as a legitimate target for a cross-border air operation, establishing a principle that Operation Sindoor would later apply at far larger scale across nine sites.

How does the LeT training course system work?

Lashkar-e-Taiba runs a graded sequence rather than a single block of instruction. It begins with a religious and ideological phase conducted at seminary level, which screens and indoctrinates new arrivals. It continues with a basic course of roughly three weeks covering small-arms familiarity and physical conditioning. It culminates in an advanced course of around three months that adds explosives handling, ambush tactics, survival skills, and reconnaissance methodology, with surveillance-specific instruction for recruits assigned to study targets before an operation. The structure means a recruit is moved deliberately through stages, and the men who attacked Mumbai in 2008 passed through this full sequence. The graded structure also means the network can adjust its output to demand. In a quiet period it can hold recruits at the religious phase, and in an active one it can push more cohorts through the advanced course. The curriculum is fixed, but the pace at which recruits move through it is not.

Are the camps hidden or operating openly?

Both, depending on the facility. The heartland campuses at Muridke and Bahawalpur operated openly for decades, appearing on commercial satellite imagery and receiving visits from international journalists, because their operators did not believe they faced any military threat. Other facilities are deliberately concealed. The Sarjal complex is housed inside the building of a Primary Health Centre, and Markaz Syedna Bilal shares a structure with a cupping-therapy clinic and a registered charity. This blend of open campuses and concealed nodes is itself a feature of the network, and the willingness to operate openly is the strongest single indicator of how secure the operators long felt.

Why does Pakistan deny the camps exist?

Pakistan’s denials serve legal, diplomatic, and financial purposes. Acknowledging a deliberate state-supported instruction network would expose Islamabad to sanctions, to renewed placement on the Financial Action Task Force grey list, and to liability in international forums. The denials are made easier by the dual-use character of many facilities, which genuinely contain mosques, schools, and clinics alongside their militant functions, allowing officials to describe a struck compound as a religious or educational site. The denial is a policy choice rather than a factual claim, and it is contradicted by Pakistan’s own conduct, including the prosecution of figures it simultaneously insists run only charities.

How close to the Line of Control are the camps?

Distance to the Line of Control varies sharply and corresponds to function. The forward facilities of Pakistan-occupied Kashmir sit closest. Markaz Ahle Hadith Barnala near Bhimber lies roughly nine kilometres from the line, Markaz Abbas near Kotli about thirteen kilometres, and the Gulpur facility around thirty kilometres back. The Punjab heartland campuses sit far deeper, with Markaz Taiba and Markaz Subhan Allah hundreds of kilometres from any contested frontier. The pattern is deliberate, because a launch facility loses its value if it sits a long march from the border, while an institutional campus benefits from depth.

Can the camp network be permanently destroyed?

Not by a single operation, and the network’s design is the reason. The system is built across three geographic zones, the Punjab heartland, the Kashmir forward edge, and the Mansehra belt, so that a strike against one layer leaves the others intact. Operation Sindoor hit nine of the roughly twenty-one publicly mapped facilities, and a heartland campus can be rebuilt while a forward node in a rented building can be re-established within a season. Permanent destruction would require not only repeated strikes but the removal of the state protection that sustains the network, since the facilities exist because an intelligence apparatus chooses to maintain them. The map shows a system that has been wounded and exposed but not abolished.

Did Operation Sindoor destroy Pakistan’s training camps for good?

No. Operation Sindoor struck nine sites on May 7, 2025, out of roughly twenty-one publicly mapped facilities, and even a struck site can be rebuilt. Satellite imagery in the months after the strikes showed rubble clearance and reconstruction work at several locations, and Pakistani authorities restricted access to the damaged compounds. The operation degraded the network and removed its assumption of immunity, but it did not abolish a system built across three geographic zones with deliberate redundancy. Lasting destruction would require sustained pressure and, more fundamentally, the end of the state protection that funds the network’s reconstruction after each blow.

Which Pakistani province has the most training camps?

Punjab and the territory of Pakistan-occupied Jammu and Kashmir hold the most significant facilities. Punjab contains the institutional heartland, including Markaz Taiba at Muridke and Markaz Subhan Allah at Bahawalpur, the two campuses that anchor the entire network. Pakistan-occupied Kashmir holds the densest single cluster of sites, because its facilities exist to feed recruits across the Line of Control and proximity to that line concentrates them. The Mansehra belt of Khyber Pakhtunkhwa holds a third group of specialised facilities. The pattern is not random: Punjab dominates because it is the political and military core of the country, and the Kashmir territory dominates by count because geography demands it.

What role does the Pakistan Army play in the training camps?

Indian and Western assessments describe the army and its Inter-Services Intelligence directorate as the network’s builder and protector rather than a passive bystander. The army is held to supply serving instructors for weapons instruction at facilities such as Sawai Nala, to provide transport corridors, and to allow camps to operate beside military real estate. Markaz Abbas at Kotli sits roughly two kilometres from a regular army garrison. The relationship is generally managed through a handler system that leaves influence without leaving evidence, which is why Pakistan can maintain official deniability. The placement of facilities near cantonments is the single hardest fact for that deniability to survive.

Are the training camps the same as madrassas?

No, though the two are connected. A madrassa is a religious seminary, and most madrassas in Pakistan teach conventional religious education with no militant function. A training camp delivers military instruction. The connection is that a specific subset of seminaries, those affiliated with the proscribed groups, functions as a recruitment feeder, screening and indoctrinating young men who are later moved into the camps. The first stage of the instruction pipeline is itself seminary-based, which blurs the line at the entry point. The honest distinction is that the militant pipeline runs through a particular network of affiliated seminaries, not through the broad madrassa system as a whole.

Why has the international community not shut the camps down?

International action has consistently fallen short of enforcement. India has submitted dossiers to the United Nations, figures connected to the network have been placed on global sanctions lists after prolonged diplomatic effort, and the Financial Action Task Force grey-listed Pakistan between 2018 and 2022. Each of these steps imposed real cost, but none dismantled the facilities, because international bodies can audit finances, demand legislation, and designate individuals without being able to police buildings inside another state. Shutting the camps down would require either the Pakistani state to choose to do so or a level of external coercion the international system has not been willing to apply. The gap between designation and enforcement is why the camps have endured.