Off air

Conflict · Cross-border airstrikes

Pakistan claimed no civilian harm. UN says Afghan strikes killed children.

UNAMA assessed at least 10 civilian deaths, most of them children, in Kunar and Helmand. Pakistan claims militant targets and no collateral damage—and says its campaign will continue.

Story trackerDay 5

1 report51 outlets92 sources

Video in its original language

Transcript · LOADING

Pakistan said its airstrikes inside Afghanistan caused no collateral damage. The United Nations mission’s assessment tells a different story: at least 10 civilians killed, most of them children, and nine people injured in Kunar and Helmand provinces early on October 1, 2026. Pakistan has nevertheless announced that its campaign will continue “at full pace.” The civilian harm is independently assessed; the assurances of precision are not. 23 61

Pakistan’s Ministry of Information and Broadcasting acknowledged conducting aerial attacks inside Afghanistan and claimed an initial 22 militants killed. That figure remains independently unverified. The United Nations Assistance Mission in Afghanistan, or UNAMA, confirmed civilian casualties in both provinces, although its casualty statement did not itself name the attacking country. Together, Pakistan’s acknowledgement and the UN assessment establish more than a disagreement between governments: civilians died in attacks Pakistan says harmed none. 61 23

Key facts

  1. At least 10Civilians killed; most were children 23
  2. 9People injured in the airstrikes 23

The clearest account of what that meant on the ground came from Kunar. Mohammad Hashim, a 30-year-old resident of the bombed village, told the Associated Press that the strike occurred around 3 a.m. He pulled an injured woman from the rubble, then found the bodies of two men and four children. His account describes a rescue and a recovery, not a complete inventory of casualties across the two provinces. 1

Those six bodies are not additional deaths to add to UNAMA’s minimum toll. The reports concern the same attacks, and no victim-by-victim reconciliation is available. A separate AFP account identified a farmer of the same name in Chugam Valley who described a woman beneath rubble and a man with four children killed. The formulations are close but not identical; they cannot responsibly be merged into one quotation or used to manufacture another total. 1 79

Another resident, Idris Khan, described the victims as civilians, including women and children. In that report, the approximately 3 a.m. timing came from a local official, not Khan. Nor should the approximate time given for the Kunar attack be treated as a timestamp for every strike in Helmand and Kunar. 71

The homes and the changing count

Taliban spokesman Zabihullah Mujahid identified the affected locations as Chugam, in Kunar’s Shultan district, and Safar, in Helmand’s Garmser district. The names also appear as Chawgam or Chogam, Shaltan, and Garmsir. Chugam and Safar are areas within those districts, not separate districts themselves. Mujahid said three civilian homes were destroyed across the two provinces; no independently verified house-by-house breakdown is available. 50 11

Location reporting was not uniform. AtlasPress’s October 1 account placed Safar in Gereshk, while Khaama Press’s account that day identified Garmser. Mujahid’s specifically identified location, also reported by TOLOnews, places Safar in Garmser. Exact strike coordinates, the complete list of buildings hit and the distribution of casualties between homes remain unavailable. 8 10 50

The casualty assessment also developed during the day. Mujahid initially reported nine civilians killed, including women and children, and 11 injured. Reuters reported that earlier Afghan-government figure on October 1. The later UNAMA assessment established a minimum of 10 civilian deaths and nine injuries. The earlier account is not the current independent assessment, but the difference between the two has not been explained. 11 23

No documented formal correction establishes whether the difference reflects verification, classification, information from different locations or another cause. Combining the later death count with the earlier injury count would produce a tally neither assessment supports. No verified subsequent official revision of UNAMA’s figures was found in the October 1 reporting. 4 23

Mujahid condemned the attacks as “an aggression and crime in violation of all accepted principles.” That is the Afghan authorities’ political condemnation, not a judicial finding of criminal responsibility. The independently assessed deaths stand without needing to turn that accusation into a verdict. 50

Precision claimed, civilian protection not demonstrated

Pakistan described intelligence-based, selective attacks on militant camps and hideouts at two locations. Its ministry claimed weapons and ammunition were destroyed and identified the alleged targets using Fitna al Khwarij and Fitna al Hindustan, labels Pakistan uses for TTP-linked and BLA-linked groups. The affiliations, militant fatalities, weapons losses and targeting precision have not been independently established for these strikes. 61 22

Maximum precautions were taken and it was ensured that no collateral damage occurs.

Pakistan Ministry of Information and Broadcasting 61

Claiming precautions were taken is different from demonstrating that they protected civilians. The UN assessment contradicts the ministry’s assurance of no collateral damage. It does not, however, establish that no militants died. Equally, a claim of militant deaths cannot erase the civilians killed.

The minimum 10 civilian deaths and Pakistan’s claimed 22 militant deaths cannot be added into a verified combined toll. There is no independently reconciled accounting of everyone killed, whether the claimed categories overlap, whether militants were present at each location or what evidence supported selecting the targets. That missing accounting matters both to understanding this attack and to assessing the justification offered for another. 1 61

Amu TV, dpa and Radio Free Europe/Radio Liberty described October 1 as the third publicly reported round of Pakistani air attacks inside Afghanistan in about 10 days, following September 21 strikes reported in Kunar and Paktika and a later September round in Kandahar, Paktia and Khost. This is a sequence of reported attacks, not a comprehensive official count of every operation. Descriptions of roughly three months of relative calm refer to the period before the late-September resumption, not a pause ending on October 1. 4 14 29 6

The wider civilian cost is already substantial. UNAMA said it had documented more than 1,100 civilian deaths and injuries in Afghanistan from cross-border armed violence so far in calendar 2026. That is a combined casualty measure—not more than 1,100 deaths, and not a tally of all violence in Afghanistan. Whether the October 1 deaths and injuries are already included is unspecified; adding 19 to the threshold would invent precision the mission did not publish. 23

UNAMA called for de-escalation, dialogue, respect for international law and prevention of further civilian harm. Pakistan’s continuation announcement, meanwhile, provides no date for another operation, target list or campaign end date. The danger is not only that the casualty accounting remains unfinished. It is that further attacks may proceed while Pakistan maintains an assurance of civilian protection that this attack’s documented consequences contradict. 23 61

Sources for this report

As aired 12 lines
  1. Pakistani airstrikes killed children in civilian homes. At least ten civilians are dead and nine people injured, contradicting the assurance of no collateral damage, while Pakistan’s campaign is set to continue. What is established about who died, and what remains unknown before any further attacks?
  2. Pakistan acknowledged conducting aerial strikes inside Afghanistan. The attacks hit Kunar and Helmand overnight into the early hours of October first, twenty twenty-six. The independent civilian casualty assessment identifies both provinces but does not itself name the attacking country. Its minimum death toll is not a final count of everyone killed.
  3. In Kunar, thirty-year-old Mohammad Hashim pulled an injured woman from rubble, then found the bodies of two men and four children. The strike was around three in the morning. That approximate timing applies to this Kunar strike, not necessarily every attack across both provinces.
  4. Those six bodies must not be added to the minimum of ten civilian deaths. They concern the same attacks, and no complete victim-by-victim accounting is available. A separate description involving a farmer with the same name refers to a woman beneath rubble and a man with four children killed. Those close but different descriptions cannot be merged into one exact quotation or used to establish another total. Another resident, Idris Khan, described civilian victims, including women and children. The roughly three o’clock timing attached to that description came from a local official, not Khan. None of these descriptions inventories all casualties in both provinces.
  5. The affected areas lie in two separate provinces. The resolved Afghan official locations are Chugam, within Shultan district in Kunar, and Safar, within Garmser district in Helmand. Chugam is also written Chawgam or Chogam; Shultan appears as Shaltan, and Garmser as Garmsir. Chugam and Safar are areas, not separate districts. A conflicting location description placed Safar in Gereshk; the resolved official description places it in Garmser. Exact strike coordinates, every building hit and a house-by-house distribution of casualties remain unavailable.
  6. The civilian count changed from nine killed and eleven injured to at least ten killed and nine injured. The earlier Afghan-government figures preceded the independent assessment. The accompanying destruction claim covers three civilian homes across Kunar and Helmand, without an independently verified house-by-house breakdown. No explanation establishes whether the difference in casualties reflects verification, classification, information from different locations or something else. There is no documented formal correction explaining it, and no verified later official revision in the October first material. Taking the higher death count and the earlier injury count would create a combination neither assessment establishes. The Afghan condemnation of the strikes as aggression and a crime is a political position, not a judicial determination of criminal responsibility.
  7. The independently assessed civilian deaths and injuries contradict the assurance that no collateral damage occurred. Pakistan’s operation was presented as intelligence-based, selective strikes on militant camps and hideouts at two locations, with an initial claim of twenty-two militants killed and large quantities of weapons and ammunition destroyed. Those alleged militant deaths, weapons losses and claims of precision have not been independently established. Saying precautions were taken is different from demonstrating that they protected civilians.
  8. The alleged targets carried the labels Fitna al Khwarij and Fitna al Hindustan, terms Pakistan uses for TTP-linked and BLA-linked groups. Those affiliations have not been independently established for these strikes. The operation forms part of Ghazab lil Haq under the vision called Azm e Istehkam. Its stated justification is protecting Pakistani citizens from threats emanating from Afghanistan, alongside a characterization of Afghanistan as a hub of terrorism. Those are the operation’s arguments, not established descriptions of the people killed in the homes.
  9. Ten civilian deaths and twenty-two claimed militant deaths cannot be added into a verified combined toll. There is no independently reconciled accounting of everyone killed, whether the claimed categories overlap, whether militants were present at each location, or what evidence supported selecting those targets. Civilian casualties do not prove that no militants died. A militant death claim does not erase the civilians killed.
  10. Three rounds of air attacks inside Afghanistan have been publicly described in about ten days. The September twenty-first round identified Kunar and Paktika; a later September round identified Kandahar, Paktia and Khost. October first brought strikes in Kunar and Helmand. That sequence is not a comprehensive official count of every operation. Roughly three months of relative calm refers to the period before attacks resumed in late September, not a pause ending with these October strikes.
  11. More than eleven hundred civilian deaths and injuries have been documented in Afghanistan from cross-border armed violence so far in twenty twenty-six. That is a combined measure of people killed and injured, not eleven hundred deaths, and not a tally of every form of violence in Afghanistan. The threshold appeared before October first and again alongside this strike assessment. Whether the current ten deaths and nine injuries are already included is unspecified. Adding nineteen to that threshold would manufacture precision the published accounting does not provide.
  12. Pakistan has announced that the campaign will continue at full pace. The civilian deaths are independently established; the claimed militant toll and targeting evidence remain unverified, and the complete casualty accounting is unresolved. That answers what can be established now, but not who else died or what further operations will do to civilians. The demands for de-escalation, dialogue, respect for international law and prevention of further civilian harm stand alongside a continuation announcement that specifies no date for another operation, no target list and no campaign end date.

Sources

92 sources51 outlets2 wire copies

Source map

Follow the evidence behind this report. The map shows its connections; the links below open the sources cited in this article.

Source map loads as you read.

Revision log

  1. r1First published.