
Kenyan Twitter influencers were earning roughly $10 to $15 a day for joining three disinformation campaigns, paid over M-Pesa, when two Mozilla researchers traced their work in 2021. Some of the same accounts also promoted consumer brands as undisclosed paid work. That is the misinformation economy in two sentences: a working industry whose income comes partly from the ad budgets and influencer fees East African founders spend every month. For a small firm, brand safety is the habit of knowing where your shillings and your name end up, and acting on what you find.
Key Takeaways
- Disinformation in East Africa is a paid trade. Mozilla’s 2021 Kenya investigation found influencers paid roughly $10 to $15 for three campaigns a day through M-Pesa, and some of those accounts also promoted brands as undisclosed paid work.
- Ad money reaches misinformation through automated placement. NewsGuard and Comscore estimated in 2021 that $2.6 billion a year in programmatic advertising went to misinformation publishers, 1.68% of display spend across a sample of 7,500 sites.
- Waste and misinformation share one pipeline. The US Association of National Advertisers found in 2023 that made-for-advertising sites took 21% of programmatic impressions and about 15% of spend in its sample.
- The global safety net has thinned. GARM, the advertiser alliance that published shared brand-safety categories, shut down in August 2024 after X sued it under antitrust law.
- The Payee Sweep (funding, voice, name, response) gives a six-person firm a two-hour check on where its money and reputation travel.
- The East African evidence is mostly Kenyan and mostly from the Twitter era. For Uganda, Tanzania and Ethiopia, your own placement report is the best data you will get this year.
How misinformation became a paid trade in East Africa
Odanga Madung and Brian Obilo, then Mozilla fellows, spent months in 2021 tracing hashtag campaigns aimed at Kenyan judges, journalists and activists. Their report, Inside the Shadowy World of Disinformation for Hire in Kenya, counted 23,606 tweets and retweets from 3,742 accounts across 11 hashtags. Eight of the 11 campaigns reached Twitter’s trending section, and one influencer told the researchers: “The main goal is to go trending on Twitter.” Coordination ran through WhatsApp groups that synchronized posting times. A few influencers had graduated to retainers of about $250 a month, and owners of verified blue-check accounts rented them out and took a cut of the fee.
Read that as an operator and the structure looks familiar. There is a supply side (accounts with reach), a demand side (clients with a message), a payment rail (M-Pesa), a coordination tool (WhatsApp) and a distribution engine (the trending algorithm). It is a small business with low overheads and fast settlement, and it runs well. The researchers also wrote the line every brand owner should read twice: many of the accounts involved “promote brands, causes and political ideologies without disclosing that they are part of paid campaigns.”
The trade has grown since. The Africa Center for Strategic Studies counted 189 documented disinformation campaigns across the continent in its 2024 mapping, nearly four times the number reported in 2022, and it treats that figure as an undercount. The campaigns it maps are political, and many are attributed to foreign states, including Russia and China. The infrastructure is message-agnostic, though. Accounts that can trend a hashtag about judges can trend one about a juice brand’s ingredients. Every documented East African case I could find is political, and I say so plainly: the capacity is documented and the commercial use at that scale is still unproven.
The audience has noticed. The Reuters Institute’s Digital News Report 2025 found TikTok was the platform Kenyan respondents most associated with false and misleading information, in a sample that represents younger English speakers, a narrower group than the whole country. Richard Edelman’s summary of the 2026 edition puts trust in news on social media at 22% and in AI chatbots at 20% across the survey. My working assumption, which still lacks East African evidence, is that customers who doubt what they read in a feed carry that doubt to the advertisement sitting in the same feed.
How an ordinary ad budget ends up funding it
Much digital advertising is bought by audience. You choose the people; an algorithm chooses the page, app or video where your ad appears. That arrangement is efficient, and it means a founder can fund a website she has yet to hear of.
NewsGuard and Comscore tried to measure the flow in 2021. Cross-referencing advertising data for 7,500 sites against NewsGuard’s credibility ratings, they estimated that 1.68% of display ad spend went to publishers of misinformation, about $2.6 billion a year worldwide and $1.62 billion in the United States alone. In December 2023 the Association of National Advertisers published a study of large US advertisers’ programmatic buying. Made-for-advertising sites, pages built to harvest ad impressions with thin or recycled content, took 21% of impressions and about 15% of spend. The same study found that at most 36 cents of each programmatic dollar reached a publisher at all.
Those figures describe big American and global advertisers. The equivalent for a Kampala or Dar es Salaam ad account is still unpublished. The mechanism transfers, however, because a founder boosting a post or running display ads buys from the same auction logic, with less supervision and smaller budgets to absorb waste.
Money also flows inside the platforms. The Mozilla report noted that the agency selling Twitter ads in Kenya offered promoted trends for $3,500 a day, which placed brand messages inside the same trending feeds the paid campaigns were gaming. So a founder’s money can reach the misinformation economy by two routes: placement spend that lands beside false content, and influencer fees paid to accounts that rent themselves to disinformation on the side.
X, GARM and the case against shared blocklists
The World Federation of Advertisers set up the Global Alliance for Responsible Media (GARM) in 2019. In September 2020 it published a brand-safety floor and suitability framework that sorted content into 11, later 12, risk categories, a shared vocabulary for its member advertisers. In 2024 X sued GARM and several advertisers under antitrust law over the advertisers who had left the platform. GARM told members it lacked the money to fight the case and closed that August.
Here is X’s case as I would argue it for them. When the biggest advertisers agree a shared list of what counts as unsafe, a handful of marketing committees decide which speech can earn a living. A publisher misjudged by a category loses that income for good and has little recourse. In East Africa that worry has teeth, because “misinformation” is a label anyone with power can attach to inconvenient reporting.
I concede the core of it. A blocklist written in New York or Brussels carries someone else’s judgment, and a crude keyword filter can starve a serious newsroom covering an outbreak or an election. Outsourced brand safety is blunt, and it was built for brands with billion-dollar budgets.
What survives the critique is narrower. One firm deciding where its own money goes is the ordinary right of every buyer: one buyer, one budget, and the opposite of collusion. Brand safety for a six-person firm is a purchasing decision, not a political campaign. It asks the same question a careful buyer asks of any supplier: who gets paid when I pay, and would I hand over that payment in person?
The Payee Sweep
Consider a Kampala skincare brand (an invented case) with six staff, a monthly budget split between Meta and TikTok ads, and two micro-influencers on retainer. Her brand meets the misinformation economy in four places, so the sweep makes four passes, and each pass ends in one written number or one changed document.
Pass one: funding. Open each ad account and pull the placement breakdown for the last 30 days. Find the spend that served outside the platform’s own feeds, in partner apps and websites. List the top 20 placements by spend and open every one. Anything you would be embarrassed to sponsor goes on an exclusion list. If most of the list is unfamiliar, move that budget to placements you can name. Write down one number: the share of last month’s spend that landed on placements you could name.
Pass two: voice. For each influencer, scroll back 90 days. Look for hashtags posted by many accounts within the same hour, attack content about people or institutions, and unlabeled paid posts. The Mozilla findings place brand promotion and paid political campaigns in the same accounts, so this is due diligence on a supplier. Then add three lines to the agreement: paid posts carry a clear label; the influencer discloses any other paid campaign during the contract; you may end the contract at once if they join a coordinated campaign. Pay through a channel that leaves a record.
Pass three: name. Search your brand on Facebook, TikTok and Instagram for lookalike pages, and ask five loyal customers whether anyone has messaged them claiming to be you. Pin one post on every account listing your official handles and the only phone numbers you sell through. The deepfake and impersonation side of this pass has its own playbook in Deepfake Defense for Small Businesses in East Africa.
Pass four: response. Name one person who makes the call, within two hours, on whether a false claim about your product needs an answer. Draft a short holding message today, while things are calm. Keep the documents that prove your claims (certifications, ingredient lists, test results, supplier invoices) in one shared folder. Agree in advance which kinds of rumor you will leave alone, because answering every one spreads it further.
The sweep runs on a weekly rhythm once the first round is done. Pass one becomes a ten-minute line in the Monday marketing review, next to the numbers in the five-number dashboard. Passes two and three come up at each contract renewal. Pass four sits in the folder until the day it saves a week.
What we know, and what is still guesswork
Numbers honesty belongs in this piece more than most, since its subject is false claims. The strongest East African evidence here is one Kenyan study of Twitter from 2021, before the platform changed owners and policies. The ad-flow estimates are American or global and four or five years old. The Africa Center count tracks political campaigns, and the Reuters Institute’s Kenyan sample covers younger English speakers.
So I hold two conclusions with different confidence. High confidence: paid disinformation exists in the region, it rents the same influencer supply that brands use, and automated placement sends some share of every untended budget to pages a founder would refuse to fund. Low confidence: the size of that share in any given Ugandan, Tanzanian or Ethiopian account. Only your own placement report answers that, which is why pass one asks you to write the number down.
Paying for a good name
The secular case for brand safety is reputation management. A brand seen beside a lie loses trust, trust is equity, and equity is money. That case is sound, and most of this essay has run on it. Yet the analysis above showed that much of the misinformation economy operates out of sight: partner-app placements a founder rarely opens and influencer side jobs she seldom hears about. Reputation logic objects only to what customers might see, so on its own terms an unseen placement that converts cheaply is a good buy. The ninth commandment closes that gap: “You shall not give false testimony against your neighbor” (Exodus 20:16). Paying someone to bear false witness is a share in the act whether anyone ever sees the receipt. Proverbs 22:1 says “A good name is more desirable than great riches,” and in that reading a good name is the sum of what you funded as well as what you said. Reputation management wants the public face and the private ledger to agree, but it can only promise that while someone is watching. The commandment asks for it in the empty room as well.
Frequently Asked Questions
What is the misinformation economy?
It is the paid industry behind false and misleading content: influencers rented for hashtag campaigns, websites that earn programmatic ad revenue, and the platforms that sell ads beside both. In Kenya, Mozilla researchers documented influencers paid roughly $10 to $15 a day over M-Pesa in 2021, with some accounts also promoting brands undisclosed.
Does a small East African business really fund misinformation through ads?
It can. Automated ad placement chooses where your ad runs, and global studies estimate that a measurable share of display spend reaches misinformation and made-for-advertising sites. The share for East African small firms is still unmeasured, so the reliable answer comes from your own placement report for the last 30 days.
How do I check whether an influencer has joined paid disinformation campaigns?
Scroll back 90 days on their accounts. Look for hashtags many accounts posted within the same hour, attack content about people or institutions, and unlabeled paid posts. Then write disclosure, labeling and termination clauses into the agreement, and pay through a channel that leaves a clear record.
Is brand safety a form of censorship?
Industry-wide blocklists can overreach, and critics, including X in its antitrust suit against GARM, argue that coordinated standards punish speech. A single firm choosing where its own budget goes is a different act. It is an ordinary purchasing decision, made with its own judgment about which suppliers deserve its money.
What should I do if a false claim about my product spreads on WhatsApp?
Let one named person decide within two hours whether it needs an answer. Use a holding message drafted in advance, and point customers to your pinned official channels. Share proof from a folder you prepared earlier, such as certifications and test results. Leave minor rumors alone, because each reply can spread them further.
Next Step
Open last month’s placement report with whoever runs your marketing, walk the four passes together, and write down the one number from pass one. If the sweep shows that nobody in the firm owns these decisions, the gap is wider than marketing, and the AVODA Business Capability Assessment will show you where else the business runs on habit. Two hours spent now costs less than the week you would otherwise spend answering a rumor you helped pay for.
Related Reading
- Deepfake Defense for Small Businesses in East Africa
- Deepfakes and the Ninth Commandment: False Witness at Scale
- WhatsApp and TikTok Commerce: Africa’s Real E-Commerce Stack
- Founder Personal Brand as Distribution: The New Funnel
Sources and Evidence
- Inside the Shadowy World of Disinformation for Hire in Kenya (Mozilla Foundation, 2021): $10 to $15 for three campaigns a day via M-Pesa; 23,606 tweets from 3,742 accounts across 11 hashtags; 8 of 11 trending; $250 monthly retainers; blue-check rental; brands promoted undisclosed; $3,500 a day promoted trends.
- “Disinformation influencers” for hire, only $15 a day (Rest of World, 2021): independent reporting on the same Mozilla findings and their Twitter monetization angle.
- Mapping a Surge of Disinformation in Africa (Africa Center for Strategic Studies, 2024): campaigns surged nearly fourfold since 2022.
- New Research Tracks Chinese, Russian Disinformation Campaigns in Africa (China-Global South Project, 2024): Russian and Chinese campaigns in the Africa Center data nearly quadrupled since 2022.
- Fighting Misinformation and Disinformation to Foster Social Governance in Africa (Al Jazeera Journalism Review): reports the Africa Center’s 189 documented campaigns and its undercount caveat.
- Top brands are sending $2.6 billion to misinformation websites each year (NewsGuard and Comscore, 2021): $2.6 billion a year, 1.68% of display spend, 7,500-site sample, $1.62 billion in the US.
- ANA: 15% of programmatic ad spend is wasted on click-bait websites (Marketing Dive, 2023): made-for-advertising sites took about 15% of programmatic spend.
- ANA Provides “First Look” at In-depth Programmatic Media Transparency Study (Fiducia): made-for-advertising sites represented 21% of impressions.
- ANA Programmatic Media Supply Chain Transparency Study, December 2023: the full study behind the 21% of impressions and 36 cents to publishers figures.
- ANA Study Finds Over $20 Billion in “Efficiency Gains” Possible in Programmatic Advertising Spend (Lexology): at most 36 cents of each programmatic dollar reaches the publisher.
- WFA Is Shutting GARM Down After X Lawsuit (Adweek, August 2024): GARM founded by the WFA in 2019; closed after X’s antitrust suit because it could not fund the defense.
- Elon Musk’s X just sued a nonprofit advertising group out of existence (CNN, August 2024): GARM’s purpose and its decision to discontinue activities.
- Explaining digital content labels (PPC Land): GARM’s September 2020 brand-safety floor and suitability framework with 11, later 12, categories.
- TikTok top source of fake news in Kenya, Reuters report reveals (Nation, 2025): TikTok as the platform Kenyans most associate with misinformation in the Digital News Report 2025.
- Reuters Institute Digital News Report 2025: Kenyan sample represents younger English speakers.
- The Reuters Institute Digital News Report 2026 (Richard Edelman): 22% trust news in social media, 20% in AI chatbots.
