In May, TikTok announced that it would automatically label AI-generated content on its platform. But this isn’t the case for all versions of the app. A new report from the Mozilla Foundation and AI Forensics finds that the Lite-Save Data version of TikTok, which is targeted at poorer users, not only doesn’t label AI-generated content, but also lacks other similar safeguards.
“Labeling is a very important tactic that platforms use to provide some form of trust and safety,” said Odunga Madungu, a Mozilla fellow and co-author of the report.
For example, users of the full version of TikTok will see labels indicating that content is extreme or depicts dangerous behavior. Similarly, some content on topics like elections and health includes notifications encouraging users to access reliable information through “resource hubs” on the app.
TikTok Lite has none of these guardrails, especially at a time when deceptive AI-generated content is at stake in elections around the world, meaning users in poorer markets have less information than those in wealthier markets about what’s fake and what’s real.
Madungu questioned why, of all the features they could have removed to optimize the app, the company included one that would make the platform safer for users. “I don’t know if this was a choice or just an oversight,” he said.
“This report contains several factual errors and a fundamental misunderstanding of our approach to safety,” a TikTok spokesperson said in a statement. “The fact is that content that violates our rules is removed from TikTok Lite just like the main app, and we offer numerous safety features.” The company declined to identify specific errors.
Lite versions of apps were a way for companies to expand their market share in regions where users had to pay high data charges or could only afford less powerful phones. In 2015, Meta, then Facebook, released Facebook Lite, a stripped-down version of the app that was more compatible with 2G data networks. That same year, the company also released Free Basics, which allowed users in the southern hemisphere to access the platform and certain other websites without being charged for data usage (apps and services that meet this standard are known as “zero-rated”). At the time, the project was widely criticized, especially in India, for offering a second-class experience to poor customers.
TikTok launched its Lite version in Thailand in 2018 and quickly expanded to other markets in Southeast Asia, including Indonesia, Malaysia and the Philippines. The app, which unlike the full version of TikTok can run on 2G and 3G networks, has now been downloaded more than 1 billion times, according to data from the Google Play Store. (TikTok Lite is only available on Android phones.)
“Most users in the global south are low-income and resource-constrained,” says Payal Arora, a professor who studies inclusive AI cultures at Utrecht University. The lite version of the app helps companies get them on board, which she says is even more important than before because “in this AI-driven, AI-hungry market, data is currency.”
