The Gold That Pays for War: Inside the ADF's Self-Financing Insurgency
The ADF was supposed to weaken as the Islamic State's money dried up. Instead it learned to tax Congolese gold, cocoa and people. This SRM analysis maps the corridor where it operates, the commercial machinery that keeps it alive, and why five years of military operations have dispersed it without defeating it.
1. The field: a corridor, not a front line
The ADF's war is not fought along a front. It is run along a corridor: Manguredjipa, Mambasa, Bafwasende, running from northern North Kivu through southern Ituri into eastern Tshopo. Manguredjipa, in the Bapere sector of Lubero territory, is a mining town in dense forest whose only real road link runs some 100 kilometres east to Butembo, the commercial capital of the Grand Nord. Mambasa, in Ituri, is the crossroads where the axes to Beni, Kisangani, Isiro and Bunia meet. The UN Group of Experts documents ADF supply and rendezvous points along the RN4 and RN44 highways, positions that let mobile units observe, resupply and fall back into forest cover.
This geography decides everything else. The group works through fixed camps, mobile units and taxation points rather than held towns. Humanitarian presence along the corridor is thin and thinning: humanitarian workers suffered hundreds of recorded security incidents in eastern DRC in 2025, access to safe drinking water in Manguredjipa itself is below 40 per cent, and each new wave of attacks pushes health zones to absorb displaced families faster than partners can reach them.
2. The force in the field
The ADF's centre sits with Musa Baluku. UN reporting places his headquarters camp, Madina, for years in Ituri, with his deputy Seka Umaru's camp in western Lubero. The mobile cutting edge belongs to named field commanders: Ahmad Mahmood Hassan, known as Abwakasi, in Lubero, whose unit the Group of Experts holds responsible for more than 600 civilian deaths in 2025 alone, and "Mzee" Meya Ssebagala in Ituri, whose group killed more than 230. Finance and logistics run through Meddie Nkalubo, co-located with Baluku, whom the UN Sanctions Monitoring Team describes moving funds through mobile payments, banking channels and his own cryptocurrency wallet.
Strength estimates differ by counting method: The Sentry puts the group at 600 to 700 combatants with some 2,000 dependants in February 2026; UN member-state estimates count 1,500 to 2,000 members including roughly 600 fighters. Either way, this is a small force. Its manpower system explains its endurance: the group replenishes itself by abduction, captives becoming its labour force, its recruits and its dependants, and it taxes the population it lives among through a token system. Farmers and travellers receive jetons as proof of payment, tokens for movement, tokens such as "Amani na Upendo" for the tax on cocoa production, documented by the UN Group of Experts and by Egmont Institute research published in January 2026.
Joint FARDC and UPDF strikes have hurt the group without decapitating it. The strikes on Madina in July 2025 and February 2026 recovered more than 200 kilograms of explosives and over 100 improvised devices, but the senior leadership survived. By mid-2026, UN reporting placed Baluku and the Madina camp relocated to Dimba, in Haut-Uélé, west of the Mambasa–Djamu road. The headquarters of the ADF now sits in a province outside MONUSCO's area of operations.
3. The financing turn: from IS money to Congolese gold
The decisive change is financial. The Sentry's investigation "Gold-Backed Brutality", published on 27 October 2026 after more than 100 interviews, establishes that external Islamic State funding declined from late 2024, after US sanctions hit key facilitators, and that gold now stands first in the group's revenue order, ahead of cocoa taxation, racketeering and kidnapping. The UN Sanctions Monitoring Team reaches the same conclusion: the ADF is now largely independent of external ISIL financing.
The mechanism is commercial, not clandestine in the usual sense. In Manguredjipa, at least 30 of the 54 gold buyers still operating in March 2026 knowingly sourced supplies from mines controlled by the ADF and other armed groups. Gold moves into licensed trading houses in Butembo, then to refineries in Uganda, then to international markets, principally the United Arab Emirates. The Sentry estimates the Manguredjipa zone alone could yield the ADF 1 million to 1.6 million dollars a year, an informed estimate built on revenue shares observed in comparable settings, not an audited figure.
The regional numbers show the size of the river the ADF drinks from. The UN Group of Experts records that Uganda exported some 62 tonnes of gold worth 6.4 billion dollars in 2025, largely attributable, in the Group's words, to gold originating from the DRC, while Ituri province's officially declared gold exports the same year were 559 kilograms. Bank of Uganda data put gold exports at 8.9 billion dollars in the year to June 2026. An October 2026 Bloomberg investigation traced nearly half of Uganda's gold exports in that period to one Entebbe firm, Thaba Investments, operating from the former address of African Gold Refinery, the refinery sanctioned by the United States in 2022 over conflict gold; the alleged links to the sanctioned network are denied by those named. On the Rwandan side, the European Union listed Kigali's Gasabo Gold Refinery in March 2025 over illegal imports of gold from M23-controlled areas, proof that gold intermediaries can be designated when the political will exists.
Kidnapping completes the revenue model. The UN Group of Experts documented ransoms of 2,000 to 5,000 dollars per victim on the Komanda–Luna axis in January and February 2026. In Mambasa, the Kivu Security Tracker counted a record 544 abductions in March 2026 alone, over 400 attributed to the ADF, with ransoms reported between 2,500 and 10,000 dollars. Abduction is simultaneously revenue, recruitment and terror.
4. Why the military answer keeps failing
Operation Shujaa, the joint FARDC–UPDF offensive launched in November 2021, has now run for five years. The UN Group of Experts' verdict is blunt: efforts to neutralise the ADF or contain its geographic expansion "remained unsuccessful", and the operation shows a recurring pattern of dispersing the group further from the Ugandan border, expanding the threat to civilians rather than reducing it. Research published in January 2026 by the Congo Research Group reaches the same finding: pushed out of its old strongholds, the ADF redeployed into Ituri and northern North Kivu and made mass violence against civilians a central strategy of survival, control and recruitment.
Three structural reasons stand out. First, Kinshasa fights on two fronts and treats M23 as the primary threat; The Sentry notes the ADF deliberately intensifies attacks when attention shifts to M23, exploiting gaps the FARDC cannot cover. Second, the FARDC, stretched thin, increasingly leans on Wazalendo militias against ADF cells around Manguredjipa, importing the indiscipline and local taxation habits of those auxiliaries into the same mining zones. Third, momentum stalls: after a June 2026 command reconfiguration, the Kivu Security Tracker reported no offensives targeting the ADF and, by August, that Shujaa had still not resumed in North Kivu. A joint meeting in Bunia on 22 October 2026 agreed to intensify operations in Ituri; whether that agreement becomes sustained pressure is an open question.
The Sentry adds a harder political judgement: that Operation Shujaa increasingly overlaps with Uganda's wider strategic and economic interests in eastern DRC, along the trade corridors through which gold and other commodities move, and that civilians it interviewed frequently believe the ADF threat sustains the conditions for that presence. SRM records this as The Sentry's assessment and a documented local perception, not as an established fact. But any honest field analysis must note that the scarcity of ADF attacks on Ugandan soil since late 2023 raises the question of what, precisely, the expanding deployment is still for.
5. The civilian bill
The cost is paid in the corridor's villages. UN reporting counts at least 431 civilians killed by the ADF between mid-September 2025 and mid-March 2026, and 468 more deaths in 211 incidents between mid-March and mid-September 2026. In September 2026 alone, OCHA reported at least 77 civilians killed in suspected ADF attacks around Biakato, Bandisende and Mambasa town. Amnesty International concluded in May 2026 that ADF abuses in Beni, Lubero, Irumu and Mambasa amount to war crimes and crimes against humanity. The Sentry records that the ADF was the second-deadliest non-state armed group for civilians in the world in 2025.
Displacement follows the attacks into the very zones humanitarians use as bases. In September 2026, displacement monitors recorded 3,872 households fleeing toward the Biena, Manguredjipa and Alimbongo health zones after ADF attacks in Mambasa pushed people across into Lubero. Health facilities are targets, not collateral: a health centre in Lubero was attacked in November 2025 with patients among the dead. For humanitarian planners, the operational meaning is direct. The mining zones are not the ADF's rear area. They are its strongholds, its treasury and its recruiting ground, and access planning that treats them as peripheral will keep being surprised.
6. The northern expansion
The group's centre of gravity is moving. UN reporting documents ADF attacks increasing across Tshopo, and the relocation of Baluku's Madina camp to Dimba places the leadership in Haut-Uélé. In April and May 2026, attacks near the Durba gold mining zone in Watsa territory were described as the group's first known action in that area in fifteen years; a FARDC spokesperson confirmed an overnight assault on Mungwere on 30 April 2026 that killed ten people. By August 2026, Radio Okapi reported more than 10,000 displaced people in Watsa and Wamba who had gone over three months without food or medical aid. In September 2026, the head of MONUSCO told the Security Council that the ADF was making extensive use of abductions "including as far as Haut-Uélé, beyond the area of responsibility accorded to MONUSCO under its mandate".
SRM flags the verification limit plainly: Haut-Uélé sits outside MONUSCO's documented protection area, incident-level confirmation there is thin, and not every armed crime in Watsa's gold country is the ADF's work; recent night raids on gold traders there are explicitly unattributed in the reporting that documents them. The structural indicator, the leadership's own relocation, is nonetheless the strongest signal available, and it points north, toward the gold.
7. SRM assessment
Our assessment, in five judgements.
First, the ADF's centre of gravity is a commercial network, not a camp. Camps can be struck; Madina was, twice. What regenerates the group is the chain of buyers, transporters, trading houses, refiners and facilitators that converts taxed production into cash. The Sentry's central recommendation, investigations and designations against those intermediaries under paragraph 7(g) of UN Security Council Resolution 2293, targets the right level.
Second, the sanctions logic Washington applied this week to the M23 file applies equally here. The designation of Rwanda's defence ministry and of Zigama, the security services' bank, rests on the principle that the financial machinery of a war is a legitimate target. The EU's 2025 listing of Gasabo Gold Refinery shows the same principle already applied to gold. What is missing is its application to the intermediaries of the ADF's gold chain, where no new US designation was made in 2025 or 2026.
Third, formalisation is the only durable competitor to taxation by gunmen, and it is not yet racing in the ADF's zones. DRC Gold Trading, the state gold buyer, exported 292 of Ituri's 559 declared kilograms in 2025, and its claim of 3.5 tonnes exported nationally in 2025 sits uneasily beside the 2.83 tonnes of artisanal exports recorded nationally by the state mining registry. Its own branch data for the first half of 2025 lists no North Kivu buying office at all. Until a legal buyer pays competitively in Lubero and Mambasa, the ADF's jeton remains the corridor's most reliable receipt.
Fourth, discipline about the ADF and M23 files must hold. The Sentry found no conclusive evidence of direct operational collaboration between the ADF and M23, and UN assessments of alleged contacts between them conflict with each other. The two insurgencies feed on the same broken political economy and the same overstretched army. They are not the same organisation, and merging them analytically produces bad operations.
Fifth, the two-front problem is now the strategy problem. Every month the FARDC's best capacity is committed against M23 and the FRN, the ADF consolidates a self-financing statelet of taxation points across three provinces and expands into a fourth. A response that sequences the threats, M23 first, ADF later, guarantees that later never comes. Rebalancing, as the UN Group of Experts and The Sentry both urge, means dedicated forces and sustained pressure on the ADF file in parallel, plus the civilian strategies, dispute resolution, youth alternatives, protection of the cocoa and gold economies from taxation, that shrink the group's local base.
8. What to watch
- Shujaa's reality after Bunia: whether the 22 October agreement produces sustained operations in Ituri and Lubero, or another cycle of strikes and dispersal.
- Designations: whether the UN, US, UK or EU open the intermediary file, traders, refiners, exporters, facilitators, on the ADF gold chain.
- The Dimba signal: further ADF consolidation in Watsa and Wamba, and any FARDC response in a province with no MONUSCO presence.
- Formalisation's race: a DRC Gold Trading buying presence in the Grand Nord, and whether legal prices can compete with the smuggling premium.
- Harvest-season violence: cocoa taxation periods in Mambasa and Lubero have historically intensified coercion; the jeton system's spread is the indicator.
- The abduction economy: whether March 2026's record abduction numbers in Mambasa repeat, and what ransom flows imply for the group's cash position.
---
Analysis by the SRM Research Team · Lead analysts: Daniel Mwendanga, Conflict, Peace & Security Analyst and Researcher / SRM-Consultant, and Dr. Brown Walls, Contributing Analyst.