Skip to content

guide

Kenya disinformation: how false claims divide communities

Paid influencers, TikTok videos and AI clips have all fed ethnic tension in Kenya. What the research found, how it spreads, and how peace actors should respond.

Takwimu editorial team, Editorial team6 min read

Disinformation in Kenya mixes paid political influence with ethnic fear. In 2022, a Mozilla report identified 130 TikTok videos that pushed disinformation and stoked violent, ethnic discriminatory narratives. Earlier, a Mozilla investigation found influencers paid Sh1,000 to Sh1,500 to join three Twitter campaigns a day. The law is unsettled, so response has to rest on speed and trust. For election law and fact-checking, see the electoral library's page on election misinformation.

Key points:

  • False claims can spread fear between communities.
  • Paid accounts make small views look like public opinion.
  • AI-generated clips are a growing concern.
  • Fast, trusted local answers work best.

What is disinformation?

Disinformation is false content spread on purpose. Misinformation is false content shared by people who believe it's true. Both can do harm. In tense places, the effect matters more than the intent.

What did Mozilla find on TikTok?

Mozilla Fellow Odanga Madung identified over 130 videos with hate speech, incitement and political disinformation. They had drawn millions of views, and many broke TikTok's own rules. After reviewing the report, TikTok said it had removed several videos and suspended several accounts. That is a large reach.

It's a sample, and it can't show how much such content exists overall. It does show that this content can reach huge audiences.

What did Mozilla find on Twitter?

The Nation reported a three-month investigation showing influencers paid to tweet pre-determined hashtags to misinform the public or attack certain people. Twitter suspended just over 100 accounts from Kenya.

Global Voices later reported that one influencer ordinarily runs at least 10 accounts, and it interviewed people hired as online influencers for the 2022 election about the playbook they used.

How does it stoke ethnic tension?

False claims often take the same shapes. One community is said to be planning an attack. A candidate is said to have promised to expel a group. An old clip of violence is reposted as new. Each of them plays on memories of 2007-08.

NCIC monitoring has detected a resurgence of coded ethnic language and incitement online and offline ahead of 2027.

What about AI?

It's a real worry. An IEBC commissioner named the misuse of artificial intelligence among the most serious threats to the 2027 election. NCIC's chairperson has spoken of AI-generated clips, and said that the commission won't rely on TikTok clips alone, only formal complaints.

Treat any shocking clip with suspicion until it's been checked. Check where and when it first appeared.

What makes people believe it?

Fear, and loyalty. People share what fits what they already believe. Messages from family WhatsApp groups feel more trustworthy than news. In a crisis, silence from officials leaves a space that rumours fill.

How does the law apply?

It's unsettled. The Court of Appeal struck down sections 22 and 23 of the Computer Misuse and Cybercrimes Act in March 2026, and an appeal is before the Supreme Court. In July 2026, the High Court struck down the 2025 cyber harassment provision. See the Act's page.

Hate speech and ethnic contempt remain offences under the National Cohesion and Integration Act. The law can punish the worst cases, but it can't keep up with a viral rumour.

Why is enforcement difficult?

The Inspector General told the Senate that foreign-hosted accounts, encrypted platforms and politicised enforcement are complicating efforts. Investigators can't read closed groups, and platforms answer slowly.

How should peace actors respond?

Check first. Find the source, the date and the place. Ask local contacts what's happening on the ground.

Respond quickly if it's false and spreading. Use short messages, in local languages, through trusted people. Say what's true before you mention the false claim. Don't repeat it in a headline.

Decide when not to respond. A rumour with little reach may grow if you draw attention to it.

Who are the best messengers?

Local ones. Chiefs, clergy, teachers and radio presenters are often trusted more than a national office. NCIC and the Inter-Religious Council plan to train faith leaders and youth influencers to identify harmful content, verify information and develop counter-messaging.

What role do radio and WhatsApp play?

Radio reaches those who aren't online. WhatsApp reaches people through those they trust. Each needs its own approach: a short voice note may do more than a long statement.

What should a county do?

Name a spokesperson for fast corrections. Set up a contact list of local radio stations, clergy and ward leaders. Agree who can approve a message within an hour. Test it before the campaign.

What should media do?

Check before you publish. The Media Council guidelines say that the media have a duty to debunk myths, stereotypes and fake news, and they ask for real-time fact-checking in debates. Avoid amplifying falsehoods for clicks.

What should platforms be asked to do?

Ask for clear reporting routes, Kiswahili and local language moderation, and quick action during elections. Record what you've reported and when, so that you can show how long it took.

What can citizens do?

Pause before you share. Ask who's saying this, and how they know. Look for a second source. If it's about another community, take extra care. Report threats, and don't forward them.

How can you track it?

Keep a log of false claims: what was said, where it appeared, how far it spread, and what you did. Watch for themes rather than single posts. A record like this helps you to see what's coming next, and to show whether your response worked.

Where TI fits

Takwimu Intelligence (TI) follows public online text and news with a digital manipulation keyword set, marks each item as unverified, and groups reports of one claim together.

To see it on your own work, book a walk-through.

Takwimu Intelligence (TI) is a conflict early warning system for Kenya.

Sources

  1. 1

    TikTok videos stoke tensions ahead of Kenya vote: Mozilla report · Al Jazeera, 10 June 2022Opened 5 November 2026.

  2. 2

    Twitter suspends shadowy Kenyan influencers · Daily Nation, 11 September 2021Opened 5 November 2026.

  3. 3

    New research: disinformation on TikTok gaslights political tensions ahead of Kenya's 2022 elections · Mozilla Foundation, June 2022Opened 5 November 2026.

  4. 4

    How to use social media to win an election: influencers from Kenya's 2022 election explain · Global Voices, 2 November 2023Opened 5 November 2026.

  5. 5

    Why NCIC is moving early to keep Kenya's 2027 election from turning violent · Dawan Africa, 25 August 2026Opened 5 November 2026.

  6. 6

    From Mungiki sect to political goons: Kenya's unfinished war on criminal gangs · Daily Nation, 29 October 2026Opened 5 November 2026.

  7. 7

    NCIC to Kenyans: we will not rely on TikTok clips, only formal complaints · Citizen Digital, 29 July 2026Opened 5 November 2026.

  8. 8

    Court of Appeal nullifies sections of Cyber Crimes Act · Capital FM, 6 March 2026Opened 5 November 2026.

  9. 9

    ODPP appeals to the Supreme Court on sections 22 and 23 · Capital FM, 14 April 2026Opened 5 November 2026.

  10. 10

    High Court strikes down key provisions of Computer Misuse (Amendment) Act, 2025 · Kenyans.co.ke, July 2026Opened 5 November 2026.

  11. 11

    Kanja reports 12 joint hate speech investigations with NCIC in a year · Capital FM on allAfrica, 25 February 2026Opened 5 November 2026.

  12. 12

    Guidelines for Election Coverage · Media Council of KenyaOpened 5 November 2026.

  13. 13

    NCIC, clergy join forces to guard against election violence · The Standard, 2026Opened 5 November 2026.

More from the library