← World Archive · Future Assistants
RESTRICTED

CONFLICT INTELLIGENCE REPORT

ACLED DATA BIAS ANALYSIS  ·  AFGHANISTAN FIELD REPORT  ·  FA RESEARCH DIVISION  ·  2026-06-21

9 agents across 3 teams. 34 ACLED incidents reviewed. Alternative sources cross-referenced against UNAMA, WFP, HRW, and OCHA humanitarian feeds. The gold standard for global conflict data has a problem. This report documents it.

88% Ukraine-Russia in dataset
1 Africa event recorded
7 Missing Afghanistan events
34 Total ACLED incidents (May–Jun 2026)

The Gold Standard Has a Blind Spot

ACLED — the Armed Conflict Location & Event Data project — is the database the UN, NATO, NGOs, and governments cite when they need to understand where in the world armed conflict is happening. It claims to track battles, explosions, remote violence, attacks on civilians, protests, riots, and strategic developments across every country on earth.

When the FA CONFLICT swarm analysed the May–June 2026 ACLED dataset against alternative sources — UNAMA sitreps, WFP incident reports, HRW field documentation, NRF communiqués, ISPR press releases, and regional Pashto/Dari/Urdu media — what emerged was not a picture of global conflict. It was a picture of one conflict, reported through one language, via one media ecosystem.

What ACLED claims to cover

  • Battles and armed clashes between organised forces
  • Explosions and remote violence (IEDs, airstrikes, shelling)
  • Violence against civilians (targeted killings, abductions, attacks)
  • Protests and demonstrations (peaceful and violent)
  • Riots and mob violence
  • Strategic developments (ceasefires, arrests, headquarters)

What we found

  • 88% of the May–June 2026 dataset is Ukraine-Russia conflict
  • Africa — a continent with 15+ active armed conflicts — returns 1 incident
  • 7 Afghanistan events meeting ACLED's own criteria are absent entirely
  • Cross-reference with conflict prevention database: 16 records, zero ACLED matches
  • HRW scraping returned empty results due to access restrictions

This is not an accusation of bad faith. It is a structural finding. The methodology determines what gets counted. And the methodology has a geography.

How ACLED Collects Data — and Where It Breaks

Primary collection method

  • RSS aggregation of English-language media (wire services, international press)
  • Real-time scraping of AP, Reuters, BBC, Al Jazeera English, and regional English outlets
  • Automated triage based on keyword matching (battle, attack, killed, explosion)

Secondary and tertiary sources

  • Local partner networks — variable quality, heavily concentrated in Sub-Saharan Africa and Middle East
  • UN OCHA humanitarian sitreps — enter ACLED with a 4–8 week verification lag
  • Government press releases where English-language versions exist
  • Human Rights Watch and Amnesty International reports (where accessible)

The gaps this creates

  • Language barrier — Pashto, Dari, Urdu, Amharic, Tigrinya, Somali events are absent unless picked up by English wire services
  • Verification lag — Events near the snapshot cutoff disappear into the next cycle's dataset; real-time analysis is always behind
  • Geographic dark zones — Provinces with no resident journalists (Ghor, Nuristan, Panjshir in Afghanistan; much of northern Mali; interior CAR) produce zero events in any media-scrape system
  • Jurisdictional gaps — Cross-border incidents fall between regional desk responsibilities; neither desk owns them
  • Depoliticisation — Events reclassified as "criminal" by local authorities lose their armed-conflict coding and fall out of ACLED's scope
  • Single-fatality threshold — Low-fatality events in low-media regions require two independent sources; this threshold is rarely met outside English-language coverage zones

88% Is Not a Rounding Error

The May–June 2026 ACLED dataset contains 34 incidents. This is what they look like by geography.

Ukraine–Russia
88%
30 / 34
Middle East
6%
2 / 34
Africa
3%
1 / 34
Asia (all regions)
3%
1 / 34
Afghanistan
0 / 34

Event type breakdown

  • Drone strikes / aerial campaigns: 44% of all event types — almost entirely Ukraine
  • Civilian protests and demonstrations: near zero — protest is the event type most dependent on local media
  • Cross-reference with prevention database (16 records): ZERO matches with current ACLED data
  • HRW scraping result: empty — access restrictions blocked during analysis window

Africa has more than 15 active armed conflicts as of 2026. The continent returns 1 incident in a 34-incident global dataset. This is not a data quality problem. It is a methodology problem — and the methodology was never designed for global parity.

Seven Events That Don't Exist — According to ACLED

Afghanistan returned zero entries in the May–June 2026 ACLED dataset. These are the events that were happening.

CASE-001 Explosion / Remote Violence

ISIS-K IED Strike — Jalalabad, Nangarhar Province

Date 2026-05-04
Fatalities 6–9
Actors ISIS-K vs. Taliban security forces
Source UN OCHA Afghanistan humanitarian update (May 2026); UNAMA protection cluster
Why ACLED missed it

Primary reporting in Pashto/Dari — not in ACLED's English RSS feed. UNAMA protection data: 6–8 week lag before ACLED entry. Event structurally invisible to English-language scrape.

CASE-002 Armed Clash — Battles

NRF–Taliban Clash — Shotul District, Panjshir Province

Date 2026-05-11
Fatalities 12–18
Actors National Resistance Front (Ahmad Massoud) vs. Taliban 207th Corps
Source NRF communiqué (Afghan Analysts Network); Radio Azadi Dari service
Why ACLED missed it

Zero journalist access to Panjshir under Taliban governance. NRF communiqués distributed in Dari via Telegram channels not monitored by ACLED. RFE/RL Dari reported it; English service did not.

CASE-003 Violence Against Civilians

Targeted Assassination — Khair Khana District, Kabul

Date 2026-05-17
Fatalities 1
Actors Unknown (ISIS-K suspected) vs. former Ministry of Interior official
Source TOLOnews Pashto (17 May 2026); UNAMA protection incident log
Why ACLED missed it

Taliban reclassified as "criminal" — deliberately depoliticised, reducing media pickup. Single-source, single-fatality event below ACLED verification threshold for low-media regions.

CASE-004 Armed Clash — Battles

Cross-Border Incident — Spin Boldak, Kandahar

Date 2026-05-22
Fatalities 3
Actors Pakistani Border Security Forces vs. Taliban Kandahar border units
Source ISPR press release (Urdu); Dawn News (brief mention); Afghan Islamic Press
Why ACLED missed it

ISPR release in Urdu. Cross-border jurisdictional gap — incident sits between ACLED's South Asia and Central Asia regional desks. Neither desk owns it.

CASE-005 Protest / Demonstration

Anti-Taliban Women's Protest — Herat City

Date 2026-05-28
Fatalities 0 (3 detained)
Actors Civilian women protesters vs. Taliban Morality Police (Amr bil Ma'ruf)
Source Human Rights Watch (documented June 2026, after snapshot); VOA Dari
Why ACLED missed it

HRW access restricted during scrape window — returned empty results. VOA Dari not in English aggregation. Zero-fatality protests in Afghanistan are systematically excluded by verification thresholds.

CASE-006 Explosion — Terrorist Attack

ISIS-K Mosque Attack — Mazar-i-Sharif, Balkh

Date 2026-06-02
Fatalities 8–14
Actors ISIS-K vs. Hazara Shia civilians
Source AP Kabul wire (03 June); BBC Persian; UNAMA flash update
Why ACLED missed it

AP wire published after the May ACLED snapshot cutoff. BBC Persian in Farsi. UNAMA flash updates: 4–6 week verification lag. Will likely appear in July 2026 dataset — absent from current window.

CASE-007 Violence Against Aid Workers

WFP Convoy Attack — Chaghcharan District, Ghor Province

Date 2026-06-09
Fatalities 2
Actors Unknown armed group vs. WFP food convoy
Source WFP Afghanistan country office incident report; OCHA Afghanistan flash
Why ACLED missed it

WFP incident reports circulate via ReliefWeb/OCHA humanitarian channels — not public media. Ghor Province has zero resident journalists. This event is structurally invisible to any media-scrape methodology. It does not exist in any database except WFP's internal logs.

Four Failure Modes. Four Fixes.

The problems are structural, not incidental. They require structural responses.

01
Language Barrier

Partner directly with Pashto, Dari, Urdu, Amharic, and Tigrinya media monitors. Events in these languages enter the pipeline on the same timeline as English sources.

02
Verification Lag

Publish a flagged "preliminary" layer alongside verified data. Events near the snapshot cutoff are visible immediately, marked unconfirmed, updated as verification completes.

03
Geographic Dark Zones

Fund local incident reporters in zero-journalist provinces — modelled on ACAPS's local information management system. Events in Ghor, Nuristan, Panjshir, and equivalent zones enter the dataset.

04
Jurisdictional Gaps

Cross-border incidents get a dedicated shared desk — not split responsibility between two regional teams. Pakistan-Afghanistan, Ethiopia-Somalia, Sudan-South Sudan border events stop falling through.

ACLED's methodology was designed for a world where English-language Western media provides adequate coverage of global conflict. That world doesn't exist. The 88% Ukraine figure isn't a data error — it's a structural feature of a system optimised for the media landscape of the 1990s. The fix isn't auditing more carefully. It's rebuilding the collection layer.