Tiktok decided to liquidate the Team of Content moderators in Berlin, including the dismissal of people working in Dutch. Moderation is to be taken over by artificial intelligence algorithms and external companies – inform Belgian media, including the Flemish broadcaster VRT and “The Brussels Times”. The decision caused employee protests and was criticized by experts.
The “Trust and Safety” team in Berlin, with about 150 people, was responsible for the moderation of content containing violence, hate speech, pornography or disinformation. Employees served many languages, including the Dutch. Now their tasks will be taken over by algorithms supported by external subcontractors – confirmed the Chinese Bytedance concern, the owner of Tiktok.
The decision met with the resistance of employees. On July 17, a protest took place in Berlin, in which about 50 people participated. They kept banners with slogans: “We trained your machines – pay us what we deserve.” The German Trade Union VER.DI announced further actions if Tiktok does not withdraw from planned layoffs.
Tiktok slows down. And can replace a man? “Only man understands the context”
Trade unions and former employees warn that artificial intelligence is not able to replace human moderation. “AI can mistakenly classify even innocent content, such as rainbow flags, as politically controversial. It's dangerous,” said Sara Tegge, one of the dismissed moderators, cited by the AFP agency.
An anonymous employee of the Berlin Dutch language team added: “We often react too late to dangerous trends, because our team is too small. Replacing us by AI will only worsen the situation. The application will become less safe.”
Artificial intelligence experts also express concerns. Olivier Cauberghs, a specialist cited by VRT Radio 1, pointed out that algorithms are unable to recognize subtle forms of hate speech, e.g. hidden in emoji, codes or music. “Only a man is able to understand the context” – he emphasized. In his opinion, the decision of Tiktok is part of the global trend of reducing investment in manual moderation, which can increase the number of inappropriate content on the platform.
Tiktok puts on AI
This is not the first time Tiktok gives up local moderation teams. Last year, the platform released the entire team in Amsterdam, which had about 300 people and served content in Dutch, English and Italian languages.
Despite criticism, Tiktok maintains that most of the content that violates the regulations is already removed thanks to artificial intelligence. The company also ensures that manual moderation will be continued in Europe, but implemented by external subcontractors.
Tiktok's decision raises questions about the future of content moderation on the Internet. Is AI really ready to fully take over responsibility for users' safety? Platform experts and employees remain skeptical.
