Information manipulation: detect, understand and respond, in Africa
Coordinated social media campaigns, fake accounts, mirror sites, amplified rumours, foreign interference: information manipulation now targets African governments, companies, major projects and executives. Cabinet 86 detects these operations early, analyses how they work and builds the response, from our offices in Paris and Abidjan, across Africa.
Why information manipulation has become a top-tier risk
A manipulation operation does not look like a conventional crisis. It starts with a few anonymous accounts, a page created the day before, a video taken out of context, then amplifiers who share without checking. Within days, a mining programme or a company can find itself at the centre of a controversy. Narratives established this way are then picked up by the media, by search engines and by generative AI, which present them as facts.
African information spaces are particularly exposed: conversations concentrated on Facebook, WhatsApp, TikTok and X, a large and unevenly independent online press, active diaspora communities, and foreign players running influence campaigns there. Responding requires knowing these environments from the inside.
What we mean by information manipulation
By this we mean any coordinated, concealed action intended to deceive an audience: disinformation (deliberate falsehood), amplified misinformation (an error repeated in good faith), inauthentic account campaigns, misuse of content, impersonation of media or institutions, foreign interference operations. We always distinguish legitimate criticism, however fierce, from manipulation: only the latter calls for our response.
Our role is defensive: we help our clients spot these operations, understand them and respond with facts, sources and credible spokespeople. We do not create fake accounts, we do not buy amplification and we do not spread misleading content: these methods destroy the very reputation they claim to protect.
A typical assignment delivers a map of the operation (players, amplifiers, timeline, narratives), an assessment of its real reach and a prioritised response plan, from controlled silence to a public response.
Our services
Seven formats, from early warning to response, for governments, institutions, companies and executives in Africa.
Early detection
Continuous monitoring of social media, online press, forums and open messaging channels, with alert rules for signs of inauthenticity: recent accounts, synchronised posts, abnormal resharing. Same-day alerts.
Campaign mapping
Identification of the accounts, pages, sites and amplifiers involved, their links and their timeline. We reconstruct the architecture of the operation, from the first accounts to the amplifiers, country by country and language by language.
Narrative analysis and attribution
Reading of the narratives spread, their targets and how they vary across countries and languages; attribution hypotheses documented from open sources, with their confidence level and the evidence that supports them.
Response and counter-narratives
A prioritised response plan: facts to set straight, spokespeople, channels, timetable. We write the talking points, posts and rights of reply, and we measure their effect on the conversation. When the campaign triggers an open crisis, our crisis communication team takes over.
Media relations and fact-checking
Sharing evidence with newsrooms and fact-checkers, requests for correction, reports to platforms under their rules, and referrals to the competent authorities in liaison with your legal advisers.
Exercises and preparation
Simulation of a hostile campaign with your executive committee or your office: hour-by-hour decisions, sign-off chain, quick-reference sheets, then an update of the response plan.
Long-term tracking
Post-crisis monitoring of residual narratives, search results and generative AI answers, with periodic reviews for management or your office. This tracking draws on our online reputation and monitoring services in Africa.
Our method: open sources and human analysis, for a proportionate response
Everything starts with monitoring. We collect from platforms, online press and open messaging channels, using monitoring tools and open source intelligence (OSINT) techniques: metadata analysis, account and domain history, reverse image and video search, community mapping on Facebook, X, TikTok and LinkedIn.
Then we analyse. A wave of hostile posts is not necessarily a coordinated operation, and a coordinated operation does not always have an audience. We assess the threat against explicit criteria (coordination, inauthenticity, reach, intent) before recommending a response, which can range from documented silence to a public response.
This expertise builds on our strategic monitoring and protects the reputation of the executives and institutions targeted; it extends into communication and influence when a narrative needs to be rebuilt.
- Coordination
- Inauthenticity
- Reach
- Intent
African environments
Campaigns first spread there on Facebook, WhatsApp and TikTok, often in several languages, and rely on a fast-moving online press and on diaspora communities. The same narratives are picked up in European and North American capitals a few hours later, then in generative AI answers.
Our offices in Paris and Abidjan and our daily work on the ground allow us to tell genuine outrage from an orchestrated campaign and a reference outlet from an opportunistic site, and to know whom to speak to when a response is needed. We work in West Africa, Central Africa, the Maghreb, Madagascar and the Indian Ocean.
Deliverables and timing
Alert: a report note within the day, with initial assessment points and a recommended stance.
Campaign mapping: a 15 to 30-page report, network graph, timeline, narratives, measured reach, attribution hypotheses.
Response plan: a decision document, talking points, timetable and tracking indicators, delivered in the days following the alert.
Permanent set-up: daily monitoring, weekly update, monthly review for management or your office.
Simulation exercise: half-day or full day, tailored scenario, debrief and quick-reference sheets.
Reports and media relations: evidence file, letters to platforms and newsrooms, follow-up of the responses received.
Examples of assignments
Major mining programme in West Africa: daily monitoring, detection of hostile narratives and their amplifiers, a response set-up and explanatory content on the project’s impact.
A Maghreb state: weekly tracking of the country’s image in the media and on social networks, identification of campaigns and the players behind them, summaries for decision-makers.
Countries in Southern Africa and the Indian Ocean: mapping of civil society and online opposition voices, identification of influential accounts and pages, reading of mobilisation dynamics.
Executive targeted by online accusations: an audit of what Google, social networks and generative AI showed under their name, separating well-founded criticism from manipulated content, a response plan and follow-up.
Frequently asked questions
What is the difference between disinformation, misinformation and information manipulation?
Disinformation is a deliberate falsehood; misinformation is an error shared in good faith; information manipulation is the coordinated, concealed action that exploits either one to deceive an audience and influence its decisions. We deal with all three, with different responses.
How can you tell whether a campaign is coordinated or spontaneous?
We examine the timeline of posts, the age and behaviour of accounts, the similarity of content, the amplifiers and the real reach. A spontaneous wave has identifiable authors and varied messages; a coordinated operation shows signs of inauthenticity that we document before drawing any conclusion.
Can you say who is behind a campaign?
Sometimes. Open source attribution remains a hypothesis, stated with a confidence level and the evidence that supports it. We never present an assumption as a certainty, because a wrong attribution makes the crisis worse.
Should you always respond publicly?
No. Responding to a campaign that has no audience can give it one. We first measure the real reach, then recommend a proportionate response: monitoring, a discreet correction with the amplifiers, a public correction, action with the platforms or the authorities.
Do you work for governments and public institutions?
Yes, as well as for companies, infrastructure programmes and executives, in West Africa, Central Africa, the Maghreb and the Indian Ocean. We turn down any assignment that would involve producing or spreading misleading content.
How quickly can you step in?
An initial assessment note can be produced within the day from the information available; a full monitoring and response set-up is in place within a few days.