TikTok is expanding the machinery it uses to label, detect and explain AI-generated content, and it is pitching the effort squarely at the Middle East and North Africa. The company announced a set of AI transparency and literacy initiatives covering new educational resources, stronger labelling technology and tougher systems for catching AI-generated spam. The practical question for anyone scrolling in the UAE is simple: does this actually change what lands in your feed? On the evidence, yes — modestly, and in the right direction.
What is TikTok changing about AI content?
TikTok is adding controls and labels that let users identify AI-generated content and dial down how much of it they see. The centrepiece for regular users is Manage Topics, a feature that gives people a lever over how much AI-generated content surfaces in their For You feed. That sits alongside the labelling stack TikTok has been building: Content Credentials metadata based on the C2PA standard, creator disclosure tools, and invisible watermarking baked into content made with TikTok’s own AI tools.
The scale claim is the headline number. TikTok says it has now labelled more than three billion AI-generated videos using that combination of methods. It has also joined the Steering Committee of the Coalition for Content Provenance and Authenticity, two years after becoming the first video platform to implement C2PA Content Credentials. That gives TikTok a seat at the table where the industry standard for provenance is actually set — which matters more than another badge, because labelling only works if platforms agree on how it is read.
How does this affect UAE users?
For UAE users, the visible change is more control over AI content and clearer signals about what is synthetic. AI-generated videos should increasingly carry labels that flag them as generated or heavily edited, and Manage Topics offers a way to reduce the volume of AI-heavy material without abandoning the feed entirely. TikTok frames MENA creators as active AI users — for storytelling, education and production — so the company is trying to keep the creative upside while stopping the feed from filling with machine-made filler.
The enforcement side is where the real value sits. TikTok says it will begin testing detection systems aimed at accounts dedicated to posting AI-generated spam, with an initial focus on politics and current events, financial advice and medical information. Those are precisely the categories where a plausible-looking AI clip does the most damage, so prioritising them is the sensible call. It builds on existing integrity work: the company says it removed more than 86 million fake accounts in the first quarter of this year. TikTok’s moderation and removal work in the region has been the subject of its recent MENA safety reporting, and the AI-spam push is a continuation of that same effort rather than a separate one.
What is TikTok doing on AI literacy?
TikTok is investing in education to help users recognise AI-generated content rather than relying on labels alone. It has partnered with the National Association for Media Literacy Education and AI expert Henry Ajder on a guide for responsible AI use, and is introducing in-app resources that teach practical skills for spotting synthetic media. Content from partners including No Filtr and Raspberry Pi has generated more than 200 million views since the programme launched in November 2025.
The money behind it is not trivial: TikTok says it has committed more than US$4 million (approximately AED 14.7 million) to these literacy initiatives so far, with a stated intention to expand their reach. That is a reasonable bet, because provenance labels only help people who understand what a label means. A watermark is useless if the viewer does not know to look for it. Pairing detection and labelling with genuine literacy work is the part that gives this announcement more weight than a typical trust-and-safety press release.
