By Souheil Mouhajir
In the first week of February, Washington signed eleven bilateral critical minerals frameworks in a single day, two of them with African states, and stood up a new coalition to replace the one it had built two years earlier. In the first week of July, the Democratic Republic of the Congo's cobalt export system stopped working because one state agency had not sent another a notification. Both weeks belong to the same story, and it is not the one the briefs tell.
Ask who is winning the contest for Africa's critical minerals and the answer arrives in the grammar of great power rivalry: the United States against China, Europe trailing, each maneuvering for cobalt, lithium, graphite and rare earths. That grammar is incomplete. The decisive contest runs inside Africa, and it takes a sharper form than the usual complaint about disunity. Two African states signed with Washington on 4 February 2026. One had spent a century building the industry it was selling into. The other granted a right of first offer over assets it had not yet named. What separated them was not the skill of their negotiators or their appetite for continental solidarity. It was what each owned on arrival. The bilateral path rewards a state that already controls its value chain, and on this continent, that describes Morocco and very nearly no one else
The African Union did not lose an argument against bilateralism. It never gave its members anything worth more at the table than what they could sell alone.
Set aside first the complaint that Africa holds no position on governing mineral resources. It has held one since 2009, when the Assembly adopted the Africa Mining Vision, a blueprint built on beneficiation, local linkages and the conversion of mineral rents into industry. In February 2025, the Union updated the doctrine for the transition era, adopting Africa's Green Minerals Strategy at its 38th Ordinary Session. The strategy's sharpest move is a rhetorical one; it renames the resource. Washington and Brussels call these minerals critical, a category defined by their own supply vulnerability. The Union calls them green, a category defined by African industrialization. On the level of language it is the most confident thing the institution has done in a decade.
Beneath the language, the machinery is thin. The body meant to implement continental mineral policy, the African Minerals Development Centre, barely functions, because member states have never ratified it. Its statute, adopted in January 2016, requires fifteen ratifications to enter into force. Four have been deposited, by Guinea, Mali, Zambia and Nigeria, and the most recent on the AU's own status list dates from June 2019. The Commission spent January 2026 lobbying CEN-SAD and the National Assembly of Chad to close what it calls an eleven country gap. A continental strategy enforced by a secretariat that eleven governments have declined to authorize is an ambition on letterhead. The Centre exists to break the pit to port model the Union keeps condemning, and it can break nothing without a mandate its own members withhold. The doctrine is not the problem. The machinery is, and the machinery belongs to the states.
Into the space, an unratified secretariat leaves, the external powers move, and their interests are not symmetrical.
China is the incumbent, and its power sits in refining rather than mining. The International Energy Agency finds China the dominant refiner for nineteen of the twenty minerals it analyses, at an average market share near seventy percent, and close to ninety percent for graphite and rare earths. It has backed the position with capital, channeling by the IEA's count nearly 57 billion dollars of financial support for key minerals across some twenty emerging economies between 2000 and 2021, and the Griffith Asia Institute records Belt and Road metals and mining engagement at record levels through 2025. It has also shown it will treat processing as a lever. It curbed rare earth exports during the trade fight with Washington, and on 11 December 2025, at a meeting convened by the National Development and Reform Commission, Chinese phosphate producers reached an industry consensus to suspend exports in principle until August 2026. Monthly phosphate exports had already fallen from roughly 950,000 tonnes in March 2022 to 13,000 tonnes in March 2025. S&P Global reported in March 2026 that the suspension was holding through the war in Iran. No African export policy touches any of this. Ore is not power. Processing is power, and processing sits in China.
The United States is the challenger, and it has moved from signing deals to building architecture. At the Critical Minerals Ministerial of 4 February 2026, it concluded eleven frameworks in a day, following ten over the preceding five months and completed negotiations with seventeen more countries. Secretary of State Rubio announced FORGE as successor to the Minerals Security Partnership, alongside a nine partner private consortium called Pax Silica. Fifty four countries attended. Two African states, Guinea and Morocco, were among the signatories.
Europe is the laggard, and its position argues with itself. At the 7th African Union and European Union summit in Luanda on 24 and 25 November 2025 the joint declaration committed to supporting minerals value chains and local processing in Africa, while the EU's own Critical Raw Materials Act sets a benchmark of processing forty percent of the bloc's annual needs inside Europe by 2030. It cannot fund both at the scale either requires. Its flagship, the Lobito Corridor, is a railway carrying copper and cobalt from the Congolese and Zambian copperbelt to Angola's Atlantic coast, backed by a United States Development Finance Corporation loan and Development Bank of Southern Africa financing that reached a package near 786 million dollars at financial close in August 2026, with a further 255 million from the African Development Bank. It is faster and cleaner than the alternatives, and it is by design a better way to export raw material, which is the outcome the Union's doctrine exists to prevent.
Put the two African signatures of 4 February beside each other, and the governing pattern appears.
Morocco did what the Vision preaches, and did it before the Vision existed. OCP Group, majority owned by the state, has access to more than seventy percent of the world's phosphate rock reserves and spent a century becoming the world's largest producer of phosphate rock and phosphate fertilizers, integrated from mine through phosphoric acid to global sales, on a share of the world phosphate products market near a third. Its current investment programme runs above fourteen billion dollars across 2025 to 2027. Rabat rejoined the African Union in January 2017 and has used one integrated champion to project influence across the continent the Union is trying to assemble by treaty.
The limits of that position are worth stating precisely, because they cut against the easy version of the Moroccan story. The kingdom's first battery materials plant, opened at Jorf Lasfar on 25 June 2025 with an initial 40,000 tonnes of capacity, belongs to COBCO, a joint venture of Al Mada and the Chinese group CNGR. OCP holds no stake in it, and the plant makes nickel manganese cobalt precursor rather than the lithium iron phosphate cathode that would run on Moroccan phosphate. The LFP line is planned and not yet built. Morocco's built position is deep and genuine in phosphate chemistry, and its entry into battery materials depends on Chinese capital and Chinese technology, which is the same point about where processing sits, arriving from the other direction.
What Morocco held was still enough. It arrived in Washington owning the chemistry, a state company with global share, and a mineral the United States had just reclassified as strategic, phosphate having been added to the critical minerals list on 7 November 2025 and reinforced by a presidential proclamation of 18 February 2026 on securing supply of elemental phosphorus. The memorandum ratified a position Rabat already held. On 4 March 2026, a month after signing, Washington dismissed its appeal in the long-running countervailing duty case against OCP, having concluded that taxing the supplier it needed made no sense.
One question the memorandum leaves open underlies the rest. Western Sahara Resource Watch, a campaign group, reports that its scope reaches offshore and seabed deposits along Morocco's Atlantic coast, waters that overlap the maritime zones of Western Sahara, a non-self-governing territory. The State Department fact sheet does not spell that scope out. The legal history is not ambiguous: on 4 October 2024 the Grand Chamber of the Court of Justice of the European Union annulled the EU agreements covering the territory in cases C-778/21 P and C-798/21 P and the joined fisheries cases C-779/21 P and C-799/21 P, holding that the consent of the people of Western Sahara is a condition of validity and had not been given. A state negotiating from a built position also negotiates over the widest possible definition of what it holds.
The Democratic Republic of the Congo produces about three quarters of the world's cobalt and holds roughly fifty five percent of global reserves, six million tonnes of an estimated eleven. On 4 December 2025, with M23 holding Goma and Bukavu since early 2025 and United Nations agencies recording around seven million people internally displaced, Kinshasa declined to pool that leverage continentally and sold it bilaterally, under pressure. The Strategic Partnership Agreement was signed in Washington the same day as the peace agreements it was meant to underwrite. Under its terms, the DRC designates an evolving Strategic Asset Reserve over which United States persons hold a right of first offer, extending to minerals already marketed for export.
Implementation moved fast. In mid January 2026 Kinshasa handed Washington a shortlist covering manganese, copper and cobalt, gold and lithium, and the first Congolese copper shipped to the United States under a Gecamines and Mercuria venture on 12 January. In February the Joint Steering Committee held its inaugural meeting and the initial assets were formally designated. On 3 February the Orion Critical Mineral Consortium, led by Orion Resource Partners with the Development Finance Corporation among its members, signed a non binding memorandum of understanding for forty percent of Glencore's interests in Mutanda Mining and Kamoto Copper Company, implying an enterprise value near nine billion dollars, with the right to direct its share of production to nominated buyers. The Strategic Partnership Agreement sets five year targets for routing state marketed exports through the Sakania and Lobito corridor: half of copper, ninety percent of zinc concentrate, thirty percent of cobalt. Continental doctrine says process at home. The bilateral agreement writes shipping ratios into treaty.
The governance problems the model invites arrived on schedule. Hard obligations fell on Kinshasa, preferential fiscal terms and privileged access for American firms, while the security promises stayed vague. Washington sanctioned the Rwanda Defence Force and four senior officials on 2 March 2026, former president Joseph Kabila on 30 April, and FDLR and M23 commanders on 2 June, which is what it looks like when a signed accord refuses to translate into compliance on the ground. On 21 January 2026, according to Congolese lawyers and rights groups who filed it, a petition reached the Constitutional Court arguing that an agreement of this reach required the parliamentary vote or referendum it never received.
The endowments are comparable. What differs is a century of accumulated industrial and administrative capacity, and the patience that capacity buys. Morocco signed because it had already built the infrastructure that mitigated the agreement’s risks. The DRC signed because it had not, and because a war in its east priced delay higher than the terms on offer.
What the continental alternative would have changed in December 2025 is not the security bargain. No AU instrument could deliver what Kinshasa was buying, which was American pressure on Kigali, and three rounds of sanctions in four months suggest Washington struggled to deliver it either. The commercial terms are different. A ratified Minerals Development Centre carrying a review mandate would have given Kinshasa a procedural reason to slow the designation of the asset reserve, which is precisely what the Congolese petitioners went to court to obtain and could not get from any continental body. A working cobalt producer bloc would have made the right of first offer harder to price, because the buyer's alternative sources would have sat inside the same bloc. Neither move requires the Union to out-negotiate Washington. Both require it to exist as a party.
One test case shows whether that is reachable. The DRC and Zambia agreed in April 2022 to build a joint battery and electric vehicle value chain across their shared copperbelt, and in March 2023 the Economic Commission for Africa and Afreximbank signed the framework for transboundary special economic zones to house it. Four years on the joint zone remains at feasibility stage, its location still disputed between the two governments, and no joint precursor plant is operating. The DRC moved ahead alone, breaking ground on the Musompo precursor zone in Lualaba in March 2025. The flagship continental coalition in this sector resolved itself into a national project, which is the article's whole argument in one file.
The usual rejoinder is the wave of export bans, read as African states taking the initiative at last. It is better read as the Africa Mining Vision attempted with the wrong tool. Since 2023 at least a dozen African governments have restricted raw mineral exports. Zimbabwe restricted raw lithium ore from December 2022 and extended the measure into a full ban on raw mineral and lithium concentrate exports on 25 February 2026. Botswana began requiring miners to sell a twenty four percent stake in new concessions to local investors on 1 October 2025 where the government declined to buy it. The DRC went furthest, banning cobalt exports in February 2025 and replacing the ban in October with an annual ceiling of 96,600 tonnes for 2026 and 2027, against 2024 output near 204,000, of which 87,000 tonnes go to producers and 9,600 are held back by the regulator, ARECOMS, as a strategic reserve. As a price policy, it worked. Reuters reported on 3 July 2026 that cobalt had surged 160 percent since February 2025, to 26 dollars a pound. The grievance behind the wave is old and fair. Africa refines around three percent of the world's refined cobalt, and almost all Congolese cobalt leaves the country as intermediate oxide and hydroxide, with the refining margin taken offshore.
Then came July. ARECOMS ruled that first half allocations unused by 30 June would be forfeited, with a customs declaration deadline of 5 July. On 2 July the Chamber of Mines wrote to say that no exporter had been able to register a declaration since the first of the month, because the regulator had never sent customs the authorising notification that keeps quota linked shipments moving. One executive put sixty to seventy five percent of companies on course to miss a deadline they were administratively prevented from meeting. The most effective instrument any African state has built in this sector, the one that moved a world price, was stopped by a form one office did not send another. It is the same failure as four ratifications out of fifteen, one level down.
The structural limit is worse than the administrative one. The Vision asked for coordinated continental beneficiation. What is happening is a dozen separate national moves, each against its own ceiling. Unilateral restriction works only where a producer approaches monopoly. The DRC can move the cobalt market because it very nearly is the cobalt market. A country exporting a fungible mineral that bans its shipment watches buyers go elsewhere.
Two conditions would let the continental model bind, and the Union meets neither. The first is a market. Every serious AU document routes value addition through the African Continental Free Trade Area, and the reason is visible in the trade data: Afreximbank puts intra African trade at about 14 to 15 percent of the continent's formal trade, and more than sixty percent of it is processed and manufactured goods, against under seventeen percent of what Africa sells outside the continent. Africans already buy each other's manufactures. There is simply not enough of that trade to absorb the output the strategy imagines. The second condition is power in the literal sense. Beneficiation eats electricity, and 22.5 percent of Congolese had access to it in 2024. ARECOMS has confirmed the country holds no active cobalt refining capacity at all. The Union's industrial ambition is throttled by the Union's own infrastructure gap. Bans move the bottleneck from the mine to the border. They do not move it off the continent, and they do not touch the refineries in China.
The reshaping everyone is writing about is real. Twenty one American frameworks in five months and a ministerial drawing fifty four countries reorder external strategy around African supply. It is also the smaller reshaping. The larger one runs inward, and the minerals have become a test of whether the continental project can make itself worth joining.
On 4 February, two African states walked into the same room in Washington. One had spent a century building the capacity that let it sign from strength, and it signed alone. The other had spent a year losing cities in its east, and it signed alone too. Nothing the African Union has built would have altered either outcome, because the Union holds doctrine, language and two frameworks, and it does not hold a ratified secretariat, a continental market deep enough to absorb what it wants processed, or a grid that stays on. Strategic autonomy is not won across the table from Washington or Beijing. It is won in whether a member state reaches that table owning something. Until the Union can put that in its members' hands, Africa's minerals will keep reshaping every strategy on earth except Africa's own.
