AI Powers Disinformation Campaigns Across Africa

Anthropic’s latest threat intelligence report reveals widespread use of its Claude AI models in propaganda, political manipulation, and surveillance operations across multiple African countries. The cases—involving the Central African Republic (CAR), Kenya, Mali, and the Democratic Republic of Congo (DRC)—demonstrate how AI is being integrated into sophisticated influence campaigns.

The most extensive case involved Mali, where Claude powered a national domestic surveillance platform targeting 25 million SIM cards. A single consultant in Bamako reportedly built this system for the Agence Nationale de la Sécurité d’État (ANSE).

In the Central African Republic, Anthropic removed an account operating as the production backbone for a Russian state-aligned information operation via Radio Lengo Songo (98.9 FM). This actor coordinated with RT, Sputnik Afrique, TASS, and the Russian House in Bangui to generate pro-Russia narratives disguised as ordinary journalism—a tactic Claude was specifically instructed to facilitate.

Researchers linked this CAR operation to Politology/Africa Corps/Wagner influence networks under the control of Russia’s Foreign Intelligence Service. The actor used Claude not only for content creation but also for internal management tasks like generating contracts and performance evaluations.

In Kenya, a single operator deployed Claude to mass-produce political tweets mimicking grassroots sentiment, particularly around energy policy changes and criticism of opposition figures—all while maintaining plausible deniability.

These operations highlight growing concerns about AI’s potential for misuse in politically sensitive contexts across the African continent.