TikTok used the AI for Good Global Summit in Geneva to unveil a fresh round of measures aimed at making artificial intelligence on its platform easier to understand, harder to abuse, and safer for the people who encounter it every day.
The announcement lands alongside the platform’s Q1 2026 Community Guidelines Enforcement Report, which shows Kenya’s video removals continuing a steady climb that has run through nearly every quarter of the past year.
Together, the two updates show how TikTok is trying to manage two connected problems at once: the rapid spread of AI generated content, and the harder task of keeping a fast growing user base safe without slowing down the creativity that draws people to the app in the first place.
Fighting AI Spam Without Killing Original Creativity
AI tools have made it far easier for anyone to produce video content, and that convenience cuts both ways. While genuine creators use these tools to tell better stories, bad actors increasingly use the same technology to flood the platform with mass produced, low value content that drowns out original voices.
TikTok says it is now testing improved detection systems built specifically to catch accounts that exist mainly to pump out AI generated spam. The company removed more than 86 million fake accounts globally in the first quarter of 2026 alone, a figure it points to as evidence that its detection tools are scaling alongside the problem.
Tom Varghese, who leads AI policy for TikTok’s Global Public Policy team, said the company wants people to have context, confidence and control over how they experience AI on the platform, and that ongoing investment in detection tools and partnerships is meant to help users spot AI content and understand how it was made.
Labelling AI Content at Scale
Transparency has become the second pillar of TikTok’s approach. The platform has now labelled more than three billion videos as AI generated, using a mix of Content Credentials, creator disclosure tools and invisible watermarking. That milestone reflects a broader industry shift toward making AI involvement visible rather than assumed.
TikTok has also joined the Coalition for Content Provenance and Authenticity Steering Committee, putting it inside the room where cross industry standards for content authenticity get decided. Sitting on that committee gives TikTok a direct hand in shaping how Content Credentials technology spreads beyond its own platform.
Building AI Literacy From the Ground Up
Recognising that tools alone will not solve the problem, TikTok has partnered with the National Association for Media Literacy Education and researcher Henry Ajder to produce a practical guide for using AI responsibly. The company is also rolling out an in app AI literacy hub for users in South Africa, Nigeria and Kenya, designed to help people recognise AI generated content and understand how it shows up on their feeds.
Since launching its AI Literacy Fund in November 2025, TikTok has committed more than four million dollars to the programme and worked with local organisations including Moxi Africa, the Centre for Journalism Innovation and Development, Africa Check, Paradigm Initiative, Eveminet and Mtoto News. Combined, these partnerships have generated more than 200 million views, a sign that demand for trustworthy AI education is real and growing.
The platform is also highlighting creators using AI constructively, among them Tonnee Ndungu and Nyandia Gachago in Kenya, Olayemi Afolabi and Comfort Obiagbaoso in Nigeria, and Motso Mike and Akhil in South Africa, alongside tools like Smart Split, AI Outline and Manage Topics that give users more control over the AI content they see.
Kenya’s Video Removals Keep Rising
TikTok’s Q1 2026 Community Guidelines Enforcement Report shows the company removed 884,591 videos in Kenya between January and March for breaching its rules. Of those, 99.7% were caught and removed before anyone reported them, and 96.3% came down within 24 hours of posting.
That figure continues a trend visible across every quarter of the past year. Kenya’s removal numbers have climbed steadily since early 2025, tracking both a growing user base and TikTok’s expanding investment in automated detection.
Table: TikTok Video Removals in Kenya, Quarterly Trend 2025-2026
| Quarter | Videos Removed | Proactive Removal Rate | Removed Within 24 Hours |
|---|---|---|---|
| Q1 2025 | ~450,000 | — | — |
| Q2 2025 | 592,037 | 92.9% | 96.3% |
| Q3 2025 | ~580,000 | — | 94.6% |
| Q4 2025 | 820,552 | 99.9% | 98.4% |
| Q1 2026 | 884,591 | 99.7% | 96.3% |
TikTok reported removing roughly 450,000 videos in Kenya during the first quarter of 2025, before climbing to 592,037 videos in the second quarter, with about 92.9% taken down before being viewed and 96.3% removed within 24 hours of posting. The third quarter saw more than 580,000 Kenyan videos pulled down, with 94.6% removed within 24 hours. Removals then jumped again in the fourth quarter of 2025, reaching 820,552 videos, with 99.9% proactively removed and 98.4% taken down within a day of posting.
Account bans followed a similar pattern. TikTok banned 108,752 accounts in Kenya during the fourth quarter of 2025, with 93,704 of those linked to suspected users under the age of 13. That figure fell in Q1 2026, when TikTok removed 48,739 accounts suspected of belonging to under 13 users, a smaller number than the prior quarter but still a significant share of total enforcement activity.
The Global Picture
Kenya’s numbers sit inside a much larger global enforcement effort. TikTok removed 184,012,576 videos worldwide during Q1 2026, roughly 0.5% of everything uploaded to the platform in that period. Automated detection technology caught 178,014,154 of those videos, while 8,838,710 were reinstated after further review. The platform maintained a 99.3% proactive removal rate globally, with 94.4% of flagged videos removed within 24 hours.
Table: Global TikTok Enforcement Snapshot, Q1 2026
| Metric | Figure |
|---|---|
| Total videos removed | 184,012,576 |
| Share of total uploads | 0.5% |
| Detected via automation | 178,014,154 |
| Videos reinstated after review | 8,838,710 |
| Proactive removal rate | 99.3% |
| Removed within 24 hours | 94.4% |
| Fake accounts removed | 86 million+ |
LIVE Streaming Under Closer Watch
Livestreaming has become an increasingly important income source for creators, which makes enforcement in that space especially sensitive. TikTok interrupted 103,847 LIVE rooms in Kenya during Q1 2026 for guidelines violations. Globally, the platform took action, including warnings and demonetisation, against 58,207,389 LIVE sessions and 21,996,667 LIVE creators over monetisation policy breaches.
TikTok frames warnings as an educational step rather than a punitive one, giving creators a chance to adjust their content before facing harsher penalties under its LIVE monetisation guidelines.
AI Generated Content Faces Its Own Enforcement Track
Beyond spam detection, TikTok is treating harmful or misleading AI generated content as its own enforcement category. The platform removed 14,261 videos in Q1 2026 under its policy covering edited media and AI generated content, a rule that requires creators to label realistic AI generated material rather than leave viewers guessing.
Labelling now happens through a combination of creator disclosure tools and automated detection models, underpinned by C2PA Content Credentials technology. TikTok’s recent move to join the C2PA Steering Committee builds directly on this enforcement track, giving the company a stake in how labelling standards evolve across the wider industry.
A Platform Trying to Scale Trust Alongside Growth
TikTok’s dual announcement, sweeping AI safety initiatives on one hand and a steadily rising Kenya enforcement report on the other, points to a platform trying to stay ahead of two fast moving trends at once. AI adoption is accelerating creativity and making moderation more complex, while a growing user base in markets like Kenya is testing the limits of automated detection at scale.
The company says it combines automated tools with thousands of trust and safety professionals worldwide to enforce its Community Guidelines consistently. Whether that combination can keep pace with both AI’s expanding role and Kenya’s rising content volumes will likely shape how much users come to trust the platform’s next report.
