Why Doesn't Sudan's War End?

Part one of two. This week: how the war is fought and funded. Next week: what it has built, who is paying, and what the Security Council decided.

For five months this newsletter has pointed its light at one strait. Sudan has appeared in nearly every issue since March 2026, almost always as a single sentence near the end, the one that begins "and meanwhile." We have never once asked the question that sentence was avoiding. Why does the largest war in the world not end?

We do not know. We cannot see inside the decisions of the two generals who run it, and we will not pretend to. What we can do is what we have done with every other war in these pages: set aside what the parties say about themselves, and follow what can be counted. Where the money comes from. Where the weapons come from. How the two sides differ, and how they do not. And who is being asked, this week, to do something about it, because on 12 September 2026 the United Nations arms embargo on Darfur expires, and the Security Council must decide before then whether to renew it as it stands, widen it to the whole of Sudan and name the drones, or let it lapse. This week, the mechanism. Next week, what the Council did with it.

The scale first, and only briefly, because in this piece the numbers are the stakes rather than the argument. Some 8.6 million people are displaced inside Sudan, by the International Organization for Migration's August 2026 count. Across its borders the UN refugee agency counts 4.6 million more, most of them refugees and asylum seekers, about a fifth of them former refugees going home to countries they once fled. Nineteen and a half million are in acute hunger, about five million of them at emergency or famine level. Thirty-three million need assistance. The United Nations' appeals to help them are between two-thirds and three-quarters unfunded. A lighthouse is only useful if it turns. This week it turns.

The ground this piece covers. Darfur, in the west, is where the embargo applies. Kordofan, in the centre, is where the siege is. Based on OCHA.

Where does the money come from?

Start with the thing both armies dig out of the ground.

Sudan is one of Africa's largest gold producers, and gold is, by a wide margin, the war's first currency. The United Nations human rights office reported in July 2026 that in 2024, the areas held by the Sudanese Armed Forces declared production of about sixty-five tonnes; some twenty-eight tonnes, worth roughly 1.6 billion dollars, left through Port Sudan as official exports; gold made up nearly half of everything Sudan sold abroad that year; and close to half of the declared output was smuggled out anyway, outside every official channel. That is the government's side of the ledger, the side that keeps books.

The Rapid Support Forces keep no books we can read. What can be observed is the ground they hold: the Darfur goldfields, Jebel Amer in North Darfur and Songo in South Darfur among them, from which the UN Panel of Experts estimated some 860 million dollars of gold was extracted in 2024 alone, an estimate that Suliman Baldo, the most experienced researcher of Sudan's gold, considers exaggerated. What can be observed is also where that gold goes. The United States Treasury, in January 2025, sanctioned the RSF's commander, Mohamed Hamdan Dagalo, and seven companies based in the United Arab Emirates that it said the RSF used to buy weapons and move money. One of them, a Dubai gold dealer trading as AZ Gold, buys Sudanese gold and carries it to Dubai; another, Capital Tap Holding, manages fifty companies across ten countries and, the Treasury said, provided the RSF with both money and military equipment.

Here is the fact this piece turns on, and it took an independent investigator to state it plainly. C4ADS, a Washington research group that traces conflict supply chains through trade records, corporate filings and satellite imagery, published a study of Sudan's mining economy in May 2025 under the title Bullion for Bullets. Its finding was not that one side profits from gold. It was that both do, that both control mines and collect taxes on them, and that the gold from both sides flows overwhelmingly to the same place. The Emirates is the primary refining hub for Sudanese gold whichever army dug it. Ninety-seven percent of the official exports from army-held Sudan in 2024 went to the UAE. As much as seventy percent of the gold smuggled out of Sudan, from RSF and army-held areas alike, is reported to arrive there too. And C4ADS noted something the sanctions lists had not: the government's own state mining companies, the Sudanese Mineral Resources Company and a firm called Sudamin, were as untouched by any sanctions list as the RSF's launderers. That stayed true for fourteen months. On 16 July this year Britain listed eleven people and companies for financing and arming the war: six tied to the RSF's Dubai network, four to the army's side, and one state miner, Sudamin, that the Foreign Office says fed both. Washington had named companies on both sides as early as June 2023, but it is, as far as we can find, the first time a sanctions notice has reached the army's own state-owned miners. Washington has listed none of them, and no one has listed the Sudanese Mineral Resources Company.

The ninety-seven percent was 2024. In 2025, after the army's government broke relations with the Emirates and suspended trade with it, the declared exports moved: Bloomberg reported in April 2026, from central bank data, that official shipments to the UAE fell from twenty-two tonnes to about eight, the Emirati share of them from 99 percent to 56, and that Egypt took nearly five tonnes. The books changed their destination. The gold that leaves without papers, by every account we have, did not.

Two figures from this summer put a size on this. In August 2026 the Financial Times traced more than a hundred million dollars in bullion, taken from Sudan's central bank in the war's opening weeks in April 2023, across Chad and South Sudan and out of Juba to the Emirates; the RSF denies involvement and the paper calls the details contested. In the same reporting an Emirati official put the Sudan-to-UAE gold trade at "just over a billion dollars" in 2025. The Sentry, an investigative group that follows war finance in Africa, had independently put the year's total at about a billion, two-thirds of it smuggled. The two figures measure different things, one the trade to one country and the other a year's whole output, so they do not confirm each other; we note only that an interested party and a watchdog land in the same range from opposite directions.

What we can observe, then, is the flow: two armies, and one buyer for everything that leaves without papers. What we cannot observe is intent. The Emirates says there is no substantiated evidence it has supported the RSF or any warring party, cites the UN Panel of Experts' own earlier language to that effect, and has said it would investigate companies named in the American sanctions. The European Union banned imports of Sudanese gold this July; the Sentry's response was that a blanket ban drives the trade further underground, behind forged certificates that say Egypt or Chad or Ethiopia, and that what is needed is pressure on the named networks and on the refiners.

There is a second currency, smaller but not small: before the war Sudan supplied seven or eight tenths of the world's gum arabic exports, the emulsifier in soft drinks, and a Sudanese expert told the Darfur news agency Dabanga this July that looted gum now ranks second only to gold in financing the war. The two commodities together are the answer to the first question. Neither army has a financing problem.

Who arms whom?

If the money is symmetrical, the weapons are not, and the difference is instructive.

The war in Sudan has become a drone war. Rosemary DiCarlo, the United Nations' political chief, told the Security Council on 24 August 2026 that drone strikes had killed some 1,100 civilians in the first six months of this year, out of about 1,400 civilian deaths the UN could verify in that period; in the first four months, drones caused more than eighty percent of them. Both sides fly them. Neither side builds them.

The most careful public accounting until this week was by ACLED, the conflict-data project, in a study this June of how Sudan's drone war has adapted to changing supply lines. On the RSF side the aircraft are Chinese: BZK-005 surveillance and strike drones, and the longer-range CH-95, made by a Chinese state contractor. They reach Darfur, ACLED found, through a supply network backed by the Emirates that has run through Libya and Chad since late 2024. The UAE denies supplying military support to either side. On the army's side the fleet comes from a different set of suppliers altogether: Iranian Mohajer-6 and Ababil-3 combat drones; Turkish Bayraktars; and the Yiha III, a Turkish-Pakistani loitering munition whose Pakistani supply line was reported halted this April by Saudi Arabia, a third country reaching in to close it.

The United Nations' own investigators reached the same ground on Thursday, and went further on the RSF side. The Independent International Fact-Finding Mission for the Sudan, the body whose hallmarks-of-genocide finding is described below, published a second report on 3 September 2026, on the drone war and the networks behind it. It finds reasonable grounds to believe that a transnational network supplied the RSF with weapons, training, logistics and foreign personnel through individuals and companies in the Emirates, Chad, south-eastern Libya and the port of Bosaso in Somalia; that the equipment, including Chinese-made howitzers, air-defence systems and CH-95 drones, was consistent with material previously exported to the Emirates; that Nyala's airport is a principal entry point; and that up to two thousand former Colombian soldiers work for the RSF, many of them operating the drones and the air defences. The Emirates told the mission that every element of this is unfounded and that its own investigation of companies under its jurisdiction found nothing; Chad said it knew of no weapons crossing its territory; Colombia acknowledged its nationals' involvement and said it opposed it. We carry all three. On the army's side the mission says its inquiry has only begun, but it has identified from wreckage a Turkish Bayraktar Akinci and the Yiha III; it names no Iranian system.

The army's Iranian channel is the one place in this section where the record is moving, and we should say so rather than smooth it. In June this year, Bloomberg reported that Sudan's military was halting new Iranian arms purchases to court American favour; in April an Iranian-born California resident had been arrested at Los Angeles airport for allegedly brokering a seventy-million-dollar deal to supply Mohajer-6 systems to Sudan's defence ministry. So we write the army's Iranian supply as historical and still alleged, not as a present-tense fact.

The tempo tells you who is exposed. In May this year the army flew seventy drone strikes across the country, seventeen of them on Nyala, the most that city has taken in any month of the war. Over the past year the army's air defences intercepted about one RSF drone in eight, the RSF's about one of the army's in forty. And the corridor itself is now a target. On 27 August 2026 Chad said a Sudanese army strike had hit a fuel market near its border where, in a Chadian security official's words, the RSF refuel, killing twelve people, a week after it had accused Sudan of striking a convoy more than a hundred kilometres inside its Ennedi region; the American embassy in Khartoum condemned the army for the strikes, and Sudan's foreign ministry said its forces respected its neighbours' sovereignty and asked for an independent investigation. What we can observe is the hardware and the networks. What we cannot observe is whether the army's Iranian door is closing or merely quieter.

What are the differences?

The money is symmetrical. Almost nothing else is, and the differences run in more than one direction, so they should be listed rather than weighed.

The first is the record of crimes. On 19 February 2026 the United Nations' Independent International Fact-Finding Mission for the Sudan published a report titled Hallmarks of Genocide in El-Fasher, and presented it in fuller form to the Human Rights Council on 8 July; it is one finding, reported twice, not two. Its scope is precise. It concerns the RSF's capture of El-Fasher on 26 and 27 October 2025, at the end of the eighteen-month siege, and it found that the RSF committed at least three of the five acts the Genocide Convention names, against the Zaghawa and Fur communities of Darfur. No finding of that weight exists against the army, and we have looked. What does exist arrived this week, in the same mission's second report: that both armies' drone operations involved serious violations of the laws of war, some of which may amount to war crimes. Against the army it documents successive strikes on Ad-Da'ein Teaching Hospital in East Darfur on 20 March 2026, the first day of Eid, that killed at least seventy people and left two million without a referral hospital, and a wedding, a displacement camp, two markets and a customary court besides. Against the RSF it documents the kindergarten and hospital at Kalogi, where successive strikes killed 114 people, some sixty of them children, and the month-by-month degrading of El-Obeid's electricity, fuel and water. So the record has three tiers, not two: a genocide finding against one side, war-crimes-grade findings against both, and, beneath them, the same imported aircraft.

The second is the books. The army keeps them. Its gold leaves through a port and a central bank, its explosives arrive through a registered company, and every one of those can be watched, which is why a British sanctions notice can name its miners. The RSF's gold leaves in trucks across Chad and its weapons arrive on night flights into Nyala. A Sudanese analyst, Amgad Fareid Eltayeb, has argued in Al Majalla that this is precisely why a Sudan-wide embargo would be symmetrical on paper and asymmetrical in practice: it bites the side that can be caught. That is a prediction, not a fact, and next week's issue is where it starts to be tested.

The third is the suppliers. The RSF's chain, as the UN mission now describes it, is one network with one alleged patron, which denies everything, and foreign hands on the controls; in some units the Colombians never handed the drones over, because the skill to fly them was scarce in the RSF's ranks. The army buys from three or four states and has its own arms industry and air force. That is why the mission could name the RSF's chain end to end and only begin on the army's: one is a supply line, the other is a state's procurement.

The fourth is the seat. The army's government holds Sudan's chair at the United Nations, argues the embargo's scope in the Council, and answered none of the fact-finding mission's requests, not even the draft of its report, sent on 28 August. The RSF answered once, proposing a roundtable, and went silent. Neither cooperates; only one is treated as a state while not doing so.

The fifth is the line. One side declared a second state in July 2025, a government of its own in Nyala with its own gold to sell; the other has spent this summer moving the front against it. What that second state is, whether it can govern, and who lives under each of the two, is next week's question.

What the differences do not include is the thing this piece began with. Neither side is short of money. Neither side builds its own weapons. And both, a UN mission now finds, have flown those weapons into hospitals and markets. The mission said one more thing, in July, when it opened an urgent inquiry into El-Obeid, and it is the reason that city keeps appearing in these paragraphs: it warned that the RSF is using the same methods there that it used around El-Fasher, encirclement, attacks on the infrastructure that keeps people alive, and the throttling of access. El-Fasher fell ten months ago and is held. El-Obeid is the siege that is happening now.

What happens on 12 September?

The United Nations has kept an arms embargo on Darfur for two decades. It covers Darfur alone: not Kordofan, where the siege is, not Khartoum, not Port Sudan, and it was written before anyone imagined a war fought by aircraft ordered from abroad. It expires on 12 September 2026 unless the Security Council renews it. On 24 August 2026 the United States tabled a draft resolution that would extend the embargo to the whole of Sudan, state in plain words that unmanned aircraft and their components fall within it, and enlarge the panel of experts that monitors violations. Britain, France and Denmark back it. Russia and China, who can each veto it, do not. Russia's ambassador, Vassily Nebenzia, called the Western measures economic strangulation and rejected what he described as an attempt to draw a moral equivalence between the two sides; China's, Li Song, said sanctions must serve only to encourage peace and must respect Sudan's sovereignty. The fact-finding mission, in its report this week, told the Council in plain words that it "should extend the arms embargo throughout Sudan."

Then there is the position of Sudan's own government, and it is the most revealing of the three. Its ambassador to the United Nations, Al-Harith Idriss, rejects the extension. A Sudan-wide embargo, he argues, would put the national army on the same footing as the RSF and infringe the country's sovereignty. Read that against the second section of this piece. A Darfur-only embargo, silent on drones, reaches the RSF's Chinese aircraft in principle and the army's Iranian, Turkish and Pakistani aircraft not at all. The government that says it is fighting a genocidal militia is arguing to keep the one instrument that names its enemy's weapons and not its own. Both armies have a reason to want the embargo left exactly as it is.

What we can observe is the lineup. What we cannot observe is the outcome, and we will not guess at it. The Council can widen the embargo and name the drones. It can let the embargo lapse, or veto its widening. Or it can do the thing that councils most often do, and roll the old instrument over as it stands, Darfur-only and silent on aircraft. We opened by asking why the largest war in the world does not end. We cannot say. We can say that neither side is short of money, that neither builds its own weapons, and that the one body with the authority to name the weapons has, this week, three choices. Next Sunday, 13 September, we report which one it made.

What would prove us wrong. We have written this piece as a set of questions, and we name here what would change the answers. We publish either way.

• The embargo vote. The Darfur embargo expires 12 September 2026. If the Council extends it to all of Sudan and names drones, the claim that the world will not act on the mechanism we describe is weakened, and we will say so plainly next week. If it lapses, is vetoed, or is rolled over as it stands, Darfur-only and silent on aircraft, the instrument that already failed has been renewed, and the claim stands.
• The findings themselves. A credible independent challenge to the fact-finding mission's genocide finding would undercut a load-bearing fact above. We have found none. The mission's findings on the army's suppliers are, by its own description, preliminary; if its expanded inquiry names them, the second section is strengthened, and if it withdraws the Akinci identification, we will report that.
• The panel of experts. Its mandate runs to 12 October 2026. If its next report adds nothing to the gold and arms flows described here beyond what The Sentry, C4ADS, the Financial Times and the mission have already published, the spine of this piece is weaker than we have presented it.

Where this comes from

We did not investigate any of this ourselves. Every fact above was established by someone else, usually at some risk. What we have done is ask what their work adds up to, and hold it in one place.

Our own work, cited above

• Every issue since 23 March: the Africa line near the end, which this piece finally answers.
• 28 June and 12 July: the method, follow the money rather than the rhetoric, applied here as a method and nothing more. This piece makes no claim that Sudan's war is connected to any other.

Primary documents and official statements

• UN Independent International Fact-Finding Mission for the Sudan: Hallmarks of Genocide in El-Fasher, 19 February 2026, presented to the Human Rights Council 8 July 2026; and Fuelling Sudan's conflict: weapons, fighters and external support networks, A/HRC/63/23, advance unedited version dated 1 September and released 3 September 2026.
• UN Office of the High Commissioner for Human Rights, 15 July 2026: gold production, export and smuggling figures for 2024; gum arabic. OHCHR press release, 8 July 2026: the El-Obeid warning.
• Rosemary DiCarlo, Under-Secretary-General for Political Affairs, to the Security Council, 24 August 2026, via Radio Tamazuj.
• International Organization for Migration, Displacement Tracking Matrix, Displacement and Return Snapshot 8, 18 August 2026. UNHCR Sudan situation data as of 10 August 2026. IPC acute food insecurity analysis for Sudan; UNICEF and UNHCR 2026 appeal funding figures.
• US Treasury, Office of Foreign Assets Control: designations of 7 January 2025 (Dagalo and seven UAE companies) and 26 June 2026 (eight people and companies, both sides).
• UK Foreign, Commonwealth and Development Office: Sanctions Notice, Sudan, 16 July 2026, eleven designations. Council of the European Union: ban on Sudan-origin gold, July 2026.
• UN Security Council Resolution 2791 (2025): Panel of Experts mandate to 12 October 2026. Security Council Report, September 2026 monthly forecast: the regime's 12 September expiry.

Investigations and research

• C4ADS: Bullion for Bullets, 27 May 2025.
• The Sentry: reports on the RSF's Emirati property and corporate network, February 2026; response to the EU gold ban, 24 July 2026.
• ACLED: How has Sudan's drone war adapted to shifting supply lines?, 4 June 2026.
• Financial Times, carried by Semafor, 4 August 2026: the central bank gold and its route to the Emirates; the Emirati official's 2025 trade figure.

Reporting and argument

• Dabanga: the embargo, 25 and 26 August; gum arabic, 22 July.
• Radio Tamazuj, Sudan Tribune, France24, DW: the US draft resolution, 24 and 25 August.
• Al Jazeera, AFP, Reuters: Chad's accounts of the 20 August Ennedi strike and the 27 August fuel market strike, and the US embassy's statement, 27 and 28 August.
• Bloomberg, 2 April 2026: Sudan's 2025 gold exports by destination, from central bank data.
• UN Panel of Experts on the Sudan, S/2025/239: the 2024 Darfur gold estimate. Suliman Baldo's dissent via Darfur24 and Ayin, January 2025.
• Bloomberg, 16 June 2026: Sudan halting Iranian weapons purchases.
• Amgad Fareid Eltayeb, Al Majalla, 28 August 2026 (the case against a Sudan-wide embargo).

Interested parties, reported as such

• The United Arab Emirates: its embassy's denial of support to any warring party, and its written responses to the fact-finding mission of 24 February and 19 June 2026.
• The Government of Sudan: Ambassador Al-Harith Idriss at the Security Council.
• The Russian and Chinese missions to the United Nations, as quoted.
• An Emirati official, to the Financial Times, on the size of the gold trade.

Image

• Map of Darfur and Kordofan: based on OCHA, Sudan Reference Map, April 2023, cropped.