TikTok is eliminating its entire team of content moderators in Berlin, including those working with Dutch, and replacing them with artificial intelligence algorithms and external staff, Belgian media reported. Protests and strikes were organized in response in July.
The “Trust and Safety” team in Berlin, numbering around 150 people, was responsible for removing content containing violence, pornography, hate speech, and disinformation, among other things. Material in multiple languages, including Dutch, was moderated, according to Flemish broadcaster VRT and The Brussels Times, which produce programs in that language. Chinese company ByteDance, which owns TikTok, announced that further moderation would be carried out by external companies supported by artificial intelligence.
On July 17th of this year in Berlin, around 50 people protested against TikTok's plans, holding banners reading, “We trained your machines – pay us what we deserve.” Ver.di, the German trade union representing service sector workers, has announced further action if the company does not withdraw its planned layoffs.
Trade unions warn that AI can misclassify even innocent content—like rainbow flags—as politically controversial. “This is dangerous. You can't completely hand over moderation to algorithms,” said Sara Tegge, one of the fired moderators, as quoted by AFP.
According to VRT, TikTok's Berlin office had 11 moderators responsible for the Dutch language. One of them, quoted anonymously, warned: “We often react too late to dangerous trends because we are too small a team. Replacing us with AI will only make the situation worse. The app will become less safe.”
Experts are also critical of the decision. “AI can't recognize subtle forms of hate speech hidden in emojis, codes, or music. Only humans have that context,” AI specialist Olivier Cauberghs told VRT Radio 1. He believes the moderator layoffs are part of a global trend of platforms withdrawing investment in manual moderation, which could lead to an increase in illegal or inappropriate content.
Last year, TikTok laid off its entire content moderation team in Amsterdam, comprising approximately 300 people. The team was responsible for reviewing content in Dutch, English, and Italian. TikTok maintains that most content that violates its rules (e.g., violence) is already removed using AI. The company also assures that manual moderation will continue in Europe, but will be handled by subcontractors.
Patryk Kulpok (PAP)
pmk/ kar/
