The Architect

Sadio Camara traveled to Moscow to build Mali's security architecture. It took Kidal, fractured the junta, and killed him.

africageopolitics

In 2021, Colonel Sadio Camara made several trips to Moscow. He was Mali’s defense minister --- removed by civilian transition officials after the August 2020 coup, reinstated after the second coup Assimi Goita staged in May 2021, partly to put him back in the position. He carried an offer: Mali would pay ten million dollars a month. In return, Russia would send fighters to do what France’s Operation Barkhane had not --- suppress the jihadist insurgency that controlled more of the country than the government did.

The Wagner Group arrived in early 2022. Two thousand fighters. In November 2023, they took Kidal --- the northern stronghold that had defied government control for over a decade. It was the one moment where what the junta bought and what Russia sold were the same thing.

Then the mechanism reversed. The Sentry’s August 2025 investigation, “Mercenary Meltdown,” documented what came next: Wagner’s operations --- insufficient intelligence, insufficient air support, no accountability --- strengthened the insurgency rather than suppressing it. Civilian casualties surged 278 percent after Wagner’s arrival. Each operation that killed civilians produced JNIM recruitment. The arrangement created the condition it was hired to fix.

By the time the report was published, the failure was structural. Eight months later, on April 25, 2026, JNIM and the Azawad Liberation Front launched the largest coordinated offensive in Mali’s history. Kidal fell. Tessalit fell. Africa Corps withdrew --- negotiating its exit with the rebels, not with the authorities who had signed the contract. Camara was killed by a car bomb at his residence in Kati.

To understand why the architecture consumed its architect, compare what Russia built in Mali to what it built in the Central African Republic.

In CAR, Wagner arrived in 2018 and within four years achieved what The Sentry called “a successful testing ground” for state capture. The mechanism was direct: provide the president with security against rebel groups, and take the mines. The Ndassima gold deposit --- valued at $2.8 billion, potentially generating a billion dollars annually --- was transferred from a Canadian company to Wagner-affiliated Midas Ressources without legal basis. No output declared. No exports recorded. Gold moved through covert channels to Dubai.

CAR worked for Russia because the client was capturable. President Touadera traded sovereignty for survival. Wagner got the economic base to sustain operations independent of state payment. The loop closed: security funded by extraction, the client too dependent to object.

Mali’s junta was not capturable in the same way. Camara had built the partnership to replace French dependency, not to substitute Russian extraction. Mali’s mining sector --- internationally held concessions, established production, regulatory infrastructure --- made the CAR model of direct seizure impractical. The junta would not surrender its most profitable assets.

Russia adapted. Not mine ownership --- a joint venture. The Yadran Group broke ground on a 200-tonne gold refinery at Senou, outside Bamako, in June 2025. Mali held 62 percent. Russia held 38 percent and the processing infrastructure. Separately, Guinea’s Conakry port became the logistics gateway: Russian-Guinean companies facilitated military deliveries alongside mineral exports. When the northern collapse came, Russia did not retreat from West Africa. It rerouted.

Two parties signed the same contract. They read it differently.

Russia’s reading: presence. Geopolitical footprint, anti-France narrative, resource access through whatever mechanism the client’s constraints permit, and a logistics network that extends beyond any single client state. If the client’s territory holds, the arrangement is more comfortable. If it doesn’t, the refinery still processes gold, Conakry still moves equipment, and the narrative still plays. The Sentry documented the detail that made the reading legible: Wagner fighters demonstrated reluctance to intervene even when the capital was directly threatened, unless financial compensation was first guaranteed. A security provider that conditions intervention on payment assurances is not selling security. It is selling presence. Presence does not require winning.

The junta’s reading: security. JNIM suppressed, territory reclaimed, regime survival guaranteed by a partner without governance conditions, without human rights lectures, without the colonial residue that made French partnership politically toxic after 2020.

Neither party deceived the other. Both readings were defensible from the contract’s terms. But only one was contingent on the security outcome. Russia’s interests survive the collapse of Malian territorial control. The junta’s do not.

The architecture produced one more outcome the junta did not commission. It fractured the institution that signed it.

Camara was Moscow’s man in Bamako. Goita was the junta’s head. By August 2025, the distinction had become an almost open struggle. Goita arrested dozens of National Guard officers loyal to Camara. He launched Operation Dugukolko, recruiting 124,000 soldiers into a structure described as parallel forces beyond the defense minister’s control. The president was building an army outside his own defense ministry.

The closer Russia was to Camara, the more Goita perceived him not as a subordinate executing policy but as a competitor backed by a foreign power. The partnership designed to secure the junta was splitting it along its own command hierarchy --- another condition the architecture created rather than solved.

Sadio Camara was forty-seven. He had made the trips to Moscow himself. He had been sanctioned by the United States for what he built and seen those sanctions lifted two months before his death. He had watched Wagner take Kidal and believed the architecture was delivering.

He was not duped. The choice between French dependency and Russian partnership was not a choice between safety and danger. It was a choice between two arrangements that serve the provider’s interests before the client’s. France’s Barkhane had not suppressed JNIM either. Camara chose the one without governance conditions. It also came without a requirement that the provider’s success depend on the client’s survival --- but that distinction was not in the contract’s text. It was in its structure, visible only from inside the architecture once it was built and locked.

The Sentry documented the failure mechanism eight months before the collapse. The architecture was already consuming itself. On April 25, it consumed its architect.

Sources

- Solen