Reporting snapshot · 2 November 2026. PICSS compiles its monthly figures from open-source incident reports rather than an official register, so the totals can be revised as incidents are verified. The think tank has not published a breakdown of who was kidnapped, where they were taken or whether any were released.
What happened
Militant attacks in Pakistan fell by 37 per cent in October, reversing a three-month run of worsening violence, but the number of people kidnapped rose to its highest monthly level in a year, according to data released by the Pakistan Institute for Conflict and Security Studies (PICSS) on Sunday, 1 November.
PICSS recorded 143 militant attacks in October, compared with 226 in September, Pakistan Today reported. The attacks killed 111 people, down from 260, and injured 69, down from 242 — declines of 57 per cent and 71 per cent respectively. Dawn reported the same headline figures.
The monthly release points to improvement across most indicators. Suicide attacks fell from three in September to one, a vehicle-borne attack that caused no deaths or injuries. September’s three suicide attacks had killed 45 people and injured 138. Counter-terrorism operations killed 156 militants and arrested seven suspects in October, PICSS said, while two security-forces personnel and three civilians were killed during those operations.
Across attacks and security operations combined, PICSS counted 53 security-forces personnel, 45 civilians and five peace-committee members killed, along with 169 militants. Another 71 people were injured: 29 security-forces personnel, 31 civilians and 11 peace-committee members.
The kidnapping figure moved in the opposite direction. PICSS recorded 88 people abducted in militant incidents during October, up from 72 in September, an increase of 22 per cent and the highest monthly total in the preceding 12 months. Balochistan accounted for 70 of the 88 cases, or nearly 80 per cent.
A fall led by Balochistan and mainland KP
The October decline was concentrated in the two theatres that drove September’s surge. Balochistan recorded 71 militant attacks, down 46 per cent from 132, with deaths from those attacks falling from 136 to 44 and injuries from 63 to 37. Security forces killed 79 militants in the province during counter-terrorism operations, while three civilians were killed and two security-forces personnel were injured, according to the PICSS figures.
Mainland Khyber Pakhtunkhwa, excluding the former Federally Administered Tribal Areas, recorded 40 attacks against 58 in September, a fall of 31 per cent. Deaths from those attacks dropped from 90 to 38 and injuries from 165 to 19, an 88 per cent decline. Operations in the region killed 42 militants, and two security-forces personnel were killed.
The former FATA districts showed a smaller improvement: 30 attacks against 34, with deaths falling from 33 to 25 and injuries rising slightly from 11 to 13. Operations there killed 35 militants.
Punjab recorded two attacks and four deaths, against none in September. Sindh recorded no attacks, though operations there led to seven arrests. Islamabad, Azad Jammu and Kashmir and Gilgit-Baltistan recorded no attacks in October.
The monthly total is a sharp contrast with September, when PICSS logged what KhabarWire reported at the time as the highest monthly number of militant attacks in 12 years. The October release puts the September comparison at 226 attacks, one more than the 225 PICSS published a month earlier, a reminder that the think tank’s monthly tally is revised as incidents are verified. The September release also counted 259 deaths from militant attacks; the October comparison gives 260.
The kidnapping signal
The fall in attack numbers does not mean the threat receded evenly. Kidnapping is a lower-frequency tactic than a roadside bombing or a raid on a checkpost, but it carries a different kind of cost: it removes people from their families and communities, can be used to extract ransoms or prisoners, and is harder to count than a firefight.
PICSS’s own weekly reporting shows how quickly the picture can change within a month. In its bulletin for 23–29 October, published on 30 October, the think tank recorded 31 militant attacks, up from 21 the week before — a 47.6 per cent increase that reversed five consecutive weeks of decline. Fatalities in that week fell from 22 to 13, while injuries rose from 11 to 15. Balochistan accounted for 15 of the attacks, while Khyber Pakhtunkhwa and the former tribal districts recorded eight each. Attacks in the former FATA rose by 166.7 per cent week on week.
That late-October rebound matters for how the monthly figure should be read. A monthly average can fall even as the last week of the month deteriorates. The 143 attacks in October are fewer than September’s total, but they are not evenly spread across the month, and the weekly data suggest that the security situation at the end of October was worse than at the start.
The kidnapping numbers point to the same unevenness. Balochistan’s attack count fell by nearly half, yet the province accounted for 70 of the 88 abductions. That combination — fewer attacks, more people taken — is consistent with groups shifting towards tactics that require less direct confrontation with security forces, though PICSS has not published an assessment of motive or responsibility for the October cases.
The longer trend
The October improvement does not undo the year’s deterioration. PICSS recorded 1,376 militant attacks in the first 10 months of 2026. Those attacks and security operations together produced 3,902 deaths, including 645 security-forces personnel, 65 peace-committee members, 595 civilians and 2,597 militants, along with 1,953 injuries. Compared with the same period in 2025, attacks rose 53 per cent and the combined death toll rose 36 per cent.
The quarterly picture is also worse than the monthly one. The Center for Research and Security Studies reported in October that violence-linked fatalities rose 83 per cent in the third quarter, to 1,415, reversing declines in the first two quarters. That figure counts deaths among civilians, security personnel and militants across both attacks and state operations, so it is not directly comparable with PICSS’s monthly attack toll.
The geographic pattern still matters. Balochistan has been the centre of the year’s violence, and Khyber Pakhtunkhwa’s southern districts and the former tribal areas have remained the main frontier of cross-border militancy. The October data show both areas improving on September, but the late-month weekly figures and the kidnapping count suggest that the improvement was neither uniform nor necessarily durable.
What is still uncertain
The October release is a think-tank tally, not an audited government count. PICSS compiles incidents from open sources, and its monthly figures are revised as more information becomes available; the one-attack difference in the September comparison is an example of that revision.
The bigger unknown is what happens next. The weekly data for 23–29 October already show a rebound, and November will test whether that was a short disruption or the start of another escalation. The kidnapping figures are harder to interpret without a breakdown of the victims, the groups involved and the outcome of each case. PICSS has not said how many of the 88 people were released, how many remain missing, or whether any ransom was paid.
What is clear is that the fall in attack numbers is real within the dataset, and so is the rise in abductions. The two trends can coexist: a campaign that is killing fewer people in direct attacks may still be able to take more of them. The next monthly release, and the weekly reports before it, will show whether October was a genuine turning point or a pause inside a longer deterioration.
