Pakistan’s Drone Activity Along the LoC Is Rising Again: What Is Islamabad Testing?

1. What Happened Between 27 and 29 August

The first significant episode occurred in the Rumlidhara area of the Nowshera sector in Rajouri. Indian reporting citing official sources states that troops observed suspected aerial activity from Pakistan around midnight on 27 August. Three drones were subsequently reported to have crossed into Indian airspace. Counter-drone systems were activated and the aircraft returned across the LoC. Security forces initiated searches to determine whether anything had been dropped. The drone activity followed another potentially important event in the same general area. An Army patrol reportedly spotted a suspected militant carrying a backpack close to the fence at Rumlidhara at approximately 7.15 pm on 27 August. Troops opened fire and subsequently searched the area. Public reporting has not established whether the individual crossed the LoC, withdrew, or was unrelated to the subsequent UAV activity. A second drone episode occurred on 28 August at approximately 8.05 pm. A UAV suspected to have originated from Pakistan was detected around Behrooti village in the Balakote-Mendhar area of Poonch. Military personnel tracked it towards the Mangalnar-Dabrote axis before it returned towards Pakistan. The area falls under the operational responsibility reported for 22 Jammu and Kashmir Rifles. Searches began at first light on 29 August to determine whether the UAV had deposited weapons, narcotics or other material. As of early afternoon on 29 August, no suspicious payload had been publicly reported as recovered from the Behrooti incident. Security forces nevertheless remained on heightened alert, while separate searches continued across forested areas of Udhampur, Kathua and Rajouri following reports of suspicious ground movement.

Recent Incident Reconstruction

The table should be read with caution. The last ground-search activity has not been established as part of the same operation as the drone sightings.

2. This Is a Pattern, Not an August Anomaly

The argument that the latest flights are merely routine or accidental becomes difficult to sustain when the 2026 chronology is considered.

On 10 January, Indian security forces recovered an apparent drone-dropped arms package at Paloora in Samba containing two pistols, three magazines, 16 rounds and a grenade. On 11 January, several drones were detected across Samba, Rajouri and Poonch, including movements around Nowshera, Teryath, Kalakote, Mankote and Ramgarh. Akashvani reported that the objects entered from the Pakistani side, hovered for several minutes and returned.

By 13 January, Army Chief General Upendra Dwivedi said that at least eight drones had been detected since the preceding weekend. More importantly, he publicly articulated two possible missions. One was observation of Indian activity. The second was identifying gaps or laxity that could potentially facilitate terrorist infiltration. India raised the issue with Pakistan through the Directors General of Military Operations. Pakistan had previously rejected Indian allegations regarding cross-border drone activity.

The activity continued. On 27 February and again on 1 March, UAVs were reported near Digwar in Poonch. During the latter incident, two to three small quadcopters reportedly attempted to violate Indian airspace before being forced back. Further suspected drone activity occurred in Samba on 10 March and again on 14 March. The March 14 search came as security forces had recovered more than 3.5 kg of heroin from consignments reportedly dropped by drones in the RS Pura sector during the preceding days.

The pattern resurfaced prominently during the summer. On 27 July, suspected drones were observed at Nandpur near the International Border in Samba and at Tarkundi-Saaj near the LoC in Nowshera. Searches found no reported payload.

Only four days later, on 31 July, the Army reportedly shot down a suspected Pakistani surveillance drone near Chakla in the Khour sector. Security officials cited by NDTV said a 64 GB memory card recovered from the aircraft appeared, on preliminary examination, to contain imagery of Indian positions and strategic locations. India Today separately reported that preliminary investigation indicated a reconnaissance role. The forensic investigation was still underway at the time, so the reported content should be considered preliminary rather than definitive.

On 4 August, another Pakistani drone was reported over Shukhmal, Devindera and Tapan villages in the Bobiyan area of Kathua's Hiranagar sector. BSF personnel initiated countermeasures but the UAV, reportedly flying at altitude, returned across the border. Searches were again launched to rule out an arms or explosives drop.

Taken together, these events show persistent activity extending from the International Border in Samba and Kathua to the LoC sectors of Rajouri and Poonch. That distribution matters because it is incompatible with viewing the problem solely as narcotics trafficking or solely as tactical military reconnaissance. The drone problem increasingly appears to contain multiple overlapping missions.

3. What Is Pakistan Trying to Achieve?

ISR and battlefield reconnaissance

Assessment Probability High

ISR is currently the strongest hypothesis. The flight behaviour reported in several incidents is notable. UAVs cross or approach the border, hover briefly, move along an axis and then retreat once countermeasures are activated. This is consistent with reconnaissance, though not uniquely diagnostic of it. More importantly, there is supporting evidence from outside the latest incidents. The drone shot down near Khour on 31 July was publicly assessed by Indian security officials as a surveillance platform. General Dwivedi's January assessment similarly identified observation of Indian activity as a likely objective.

Post-Sindoor, even apparently mundane tactical reconnaissance has greater value. Operation Sindoor demonstrated that the India-Pakistan military competition had expanded decisively into the drone and counter-drone domain. Pakistan employed drones and missiles against multiple Indian military targets during the May 2025 confrontation, while India relied on an integrated Counter-UAS Grid and layered air defence to respond. Peacetime or grey-zone UAV reconnaissance therefore allows a potential adversary to maintain familiarity with the tactical environment without committing aircraft or personnel.

The intelligence value is not necessarily confined to photographing a bunker. Repeated reconnaissance can reveal changes in deployments, construction, movement patterns and the general security posture. It can also help determine which border stretches are being more intensively watched at a particular time. That alone makes recurring UAV activity strategically meaningful.

4. Infiltration Support

Assessment Probability Moderate to High

The second hypothesis deserves serious attention because the latest Rumlidhara episode included both aerial activity and reported suspicious ground movement. However, correlation cannot be converted into causation. There is currently no public evidence showing that the individual carrying a backpack was being guided by the drones or that the UAVs were transmitting information to an infiltrating group.

The hypothesis is nevertheless supported by the wider operating environment. In July 2025, officials told PTI that drones were increasingly replacing elements of traditional over-ground networks. They alleged that UAVs were being used to monitor approaching troops, provide situational awareness and assist infiltration, while in some cases even supplying food to militants operating in difficult terrain. The same reporting said Pakistani handlers were increasingly interested in drone reconnaissance ahead of infiltration attempts.

The Indian Army Chief's January 2026 comments independently reinforce this concern. Dwivedi specifically raised the possibility that UAVs were looking for security gaps through which terrorists might subsequently be sent. This produces a potentially important evolution in infiltration methodology. Historically, infiltration required substantial human preparation. Guides had to understand patrol patterns, terrain, fencing and safe routes. Every additional human participant created exposure to interception, interrogation and counter-intelligence penetration. A UAV can perform part of that reconnaissance without exposing a human scout. It does not eliminate the need for guides, facilitators or local support, but it can reduce dependence on them.

5. Arms, Ammunition and Logistics

Assessment Probability Moderate

Drone logistics are not theoretical in the Jammu theatre. They have already been demonstrated. The January 2026 Samba recovery provides a clear contemporary example. A suspected Pakistani drone crossed the border and an arms package containing pistols, ammunition and a grenade was recovered shortly afterwards.

The wider western frontier has an even larger drone-smuggling problem. BSF reported recovering 278 rogue drones during 2025, while also seizing more than 380 kg of heroin and over 200 weapons along India's western border. These figures cover the wider Pakistan-facing frontier and cannot be treated as Jammu-only statistics, but they illustrate the scale to which unmanned systems have become embedded in cross-border trafficking. The Indian government has consequently placed considerable emphasis on expanding anti-drone capabilities. In May 2026, Home Minister Amit Shah said anti-drone systems would begin to be installed along the border within six months while stressing that authorities must also identify the networks receiving drone-borne consignments after they land.

For the latest Behrooti and Rumlidhara incidents, however, no weapons or narcotics drop has yet been publicly confirmed. That distinction matters. The fact that security forces conduct immediate ground searches after a drone sighting demonstrates the seriousness with which the logistics threat is treated. It does not prove that every drone is carrying cargo.

6. Are the Drones Testing Indian Counter-UAS Defences?

Assessment Probability Moderate, Confidence Low to Moderate

This is the most interesting hypothesis, but also the easiest to overstate. Repeated penetrations can theoretically generate useful information regarding whether a UAV is detected, approximately when a response begins, and whether particular methods are sufficient to force it to withdraw. The recurrence of short-duration incursions followed by rapid withdrawal after Indian countermeasures is therefore noteworthy. There is, however, no public evidence demonstrating that Pakistan is deliberately conducting a structured counter-UAS mapping campaign. The same behaviour can be explained more simply. A reconnaissance or smuggling drone that encounters resistance naturally retreats. CNAWS therefore assesses counter-UAS probing as plausible but not established.

What is more certain is that every repeated interaction between drone and counter-drone system produces learning on both sides. India learns more about the signatures, routes and behaviour of incoming aircraft. Operators across the LoC can likewise observe whether a sortie succeeds, fails or triggers an immediate response. The competition is consequently iterative even without a formally designed testing programme.

7. The Post-Sindoor Change

Operation Sindoor altered the psychological and operational context surrounding every subsequent drone intrusion. During the May 2025 confrontation, Pakistan attempted to engage numerous Indian military targets using drones and missiles. India's official account states that the Integrated Counter-UAS Grid and air-defence systems intercepted those attempts. Pakistan and India then moved back from open conflict, but the technology demonstrated during those four days did not disappear when the firing stopped.

What appears to have emerged in 2026 is a lower-intensity application of the same unmanned domain. Rather than large-scale strike packages, Indian forces are repeatedly encountering individual UAVs or small groups. Rather than overt combat, the missions appear centred on surveillance, border penetration and potential delivery. This represents a form of persistent grey-zone friction. The economics strongly favour the initiator. A small UAV is comparatively expendable. Its detection can still require surveillance, counter-UAS activation, ground searches and manpower commitments. Even an unsuccessful drone therefore generates a security response.

For India, the problem is not simply intercepting an aircraft. It is distinguishing, often within minutes, between a harmless observation platform, a weapons carrier, an infiltration-support asset and a potential attack system. That ambiguity is itself operationally useful to an adversary.

8. The Geography Matters

The most important geographical feature of the 2026 pattern is its spread. Samba and Kathua offer a different operating environment from the rugged LoC sectors of Rajouri and Poonch. Yet UAV activity has repeatedly appeared across both. This suggests that the drone ecosystem should not be analysed as one homogeneous mission set. Along the International Border, repeated recoveries of weapons and narcotics make logistics a particularly important concern. Along the LoC, especially in Rajouri and Poonch, the combination of difficult terrain, infiltration concerns and militant activity increases the potential value of reconnaissance and route support. The July 27 sightings are particularly illustrative because suspected UAV activity occurred near Nandpur in Samba and Tarkundi-Saaj in the Nowshera sector during the same reporting period. The August incidents now extend that continuity into Rumlidhara and Behrooti. This is why the latest event should not be reduced to a headline stating that "a Pakistani drone was spotted over Poonch." The operational picture is wider.

9. CNAWS Assessment Matrix

10. Outlook

CNAWS assesses that Pakistani-origin drone activity along the Jammu frontier is likely to continue during the next three to six months. The technology provides too many advantages to disappear from the operating environment. UAVs can conduct reconnaissance, move small payloads, support infiltration, reduce dependence on human couriers and impose repeated security costs at relatively low risk.

The most likely near-term trajectory is therefore not a dramatic drone attack but continued persistent, low-signature incursions punctuated by occasional logistics drops and reconnaissance missions. The more dangerous development would be integration.

A future infiltration package could combine unmanned reconnaissance, a small logistics drop, ground infiltrators, remote handlers and pre-positioned support. No public evidence demonstrates that such a fully integrated system was employed at Rumlidhara or Behrooti this week. The current pattern nevertheless shows several of the individual building blocks already exist. That distinction should guide the assessment.

The August incidents are not yet proof of a new infiltration architecture. They are, however, credible warning indicators that the LoC's unmanned dimension is becoming more persistent, geographically distributed and operationally diverse.

Bottom Line

The important question after the latest sightings is therefore not whether India successfully forced four suspected drones back across the border. It is what those drones were trying to learn or enable before they turned around. The available evidence points most strongly towards reconnaissance. The documented history of weapons drops keeps logistics firmly in the assessment. The coincidence of UAV activity with suspicious movement near the LoC makes infiltration support impossible to dismiss. Counter-UAS probing remains plausible, although unproven. What is increasingly clear is that the drone is no longer an occasional addition to cross-border activity. It is becoming an enduring instrument of the India-Pakistan contest below the threshold of open war.