Hello,

New research by Mozilla and AI Forensics exposes how the app “Tiktok Lite - ‘Save Data’”, a stripped-down version of TikTok downloaded 1 billion times, is missing critical safety features.

Big tech companies routinely introduce ‘lite’ versions of their products that are designed for regions with slow internet connections, primarily in Global Majority countries. But our new research has revealed that TikTok Lite - “Save Data” doesn’t just reduce the required bandwidth – it is also missing basic safety features like labels for videos containing graphic images, dangerous stunts and AI-generated content, as well as the options to filter unwanted or harmful content. By failing to include these critical user protections, TikTok Lite - “Save Data” is potentially dangerous in the context of elections, public health and in its dissemination of harmful disinformation.

Many technology platforms have a history of neglecting users outside of the EU and US, where there is currently less strict regulation and enforcement.

Unless people stand up to them to demand safer products, they have no reason to change. Mozilla and AI Forensics’s new research reveals the exact differences between TikTok’s Lite - “Save Data” and “classic” TikTok apps - and with your support, the Mozilla community can demand that TikTok affords all users the same protections.

Tell TikTok that all their users deserve to be safe – and call for a revised TikTok Lite - “Save Data” app that offers the same protection features as the TikTok app.

When placed side-by-side, the disparities in the TikTok’s implementation of trust and safety measures become glaringly obvious. Here’s a few examples:

Labelling of AI-generated content – Content generated by AI has the potential to mislead viewers and exacerbate misinformation, particularly during pivotal moments such as elections, a risk that TikTok itself has publicly acknowledged.¹ In its main app, TikTok has introduced a label to inform users when a content is AI-generated or has been significantly modified using AI. But on TikTok Lite - “Save Data”, this label is yet to be introduced, leaving users more vulnerable to AI-generated content. TikTok committed earlier this year to label AI-generated content across its platform.² What reasons would they have for not following through on this promise to people in their ecosystem who choose to use a ‘lighter’ version of their app?

AI Generated Content Labelling

Labelling (highlighted in the red box) is only present on TikTok (full)

Dangerous Act Labelling – TikTok has faced serious scrutiny for sharing videos of risky challenges, which others have then attempted to replicate in the offline world - sometimes with deadly consequences.³ To counter this, in December 2020, TikTok introduced a ‘dangerous acts label’ to relevant videos, cautioning viewers with a disclaimer stating “The actions in this video are performed by professionals or supervised by professionals. Do not attempt.”, or “Participating in this activity could result in you or others getting hurt.” ⁴ While this feature is available on the main TikTok app, it is notably absent from TikTok Lite - “Save Data”. Why should TikTok omit such a critical feature from such a large part of its ecosystem?

“Do not perform the actions in this video” also not present on TK Lite

Labelling (highlighted in the red box) is only present on TikTok (full)

User controls – Having control over your experience on an app is fundamental, as it allows users to foster a safe and personalised online experience. However, TikTok Lite - “Save Data” lacks critical user controls which were introduced to reduce potential harms on their app, like the ability to filter spam and offensive comments, and a “restricted mode”, which lets users of the full app limit their exposure to unwanted content.

Content preference settings are not available on TikTok Lite.

But these three examples only scratch the surface – Mozilla and AI Forensics’s latest research has revealed a pattern of scaled-back or absent safety features on TikTok Lite - “Save Data”. This trend mirrors a well-established pattern among big tech companies, releasing substandard and unsafe products Global Majority countries - despite the fact these countries often make up the majority of their user base.

And it’s a trend that builds on Mozilla’s previous research on this topic, where big tech companies, like TikTok, have markedly fewer policies in place to protect users in Global Majority countries.⁵ To make sure the internet is a global resource and safe space for everyone, we are now calling on TikTok to update their TikTok Lite - “Save Data” app and offer the same protection to all its users.

Sign Mozilla’s petition and tell TikTok to stop sacrificing the safety and welfare of their users in the Global Majority. TikTok must update their TikTok Lite - “Save Data” app to ensure that safety of all TikTok users are prioritised equally.

Thank you for everything you do for the internet.

Nicholas Piachaud
Director, Campaigns
Mozilla

More Information:

1. TikTok. About AI-generated content.

2. TechCrunch. TikTok will automatically label AI-generated content created on platforms like DALL·E 3. 9 May 2024.

3. Independent. TikTok ‘put children in danger’ by failing to take down ‘challenge’ videos, claim grieving parents. 1 March 2024.

4. TechCrunch. TikTok expands Community Guidelines, rolls out new ‘well-being’ features. 15 December 2020.

5. Mozilla Foundation. Platforms, Promises and Politics. 27 February 2024.