AI Takes Over Moderation As TikTok Lays Off German Team
TikTok's decision to replace its entire German trust and safety team with AI and outsourced workers has triggered strikes and widespread concern over platform safety.
Strikes in Berlin Over Mass Layoffs
Tensions are high in Berlin as TikTok workers are striking to protest the mass layoff of the company’s entire trust and safety team. The social media giant plans to eliminate 150 jobs in its Berlin hub, effectively dismantling the team responsible for removing harmful content for its 32 million German-speaking users. The company intends to shift these critical responsibilities to a combination of artificial intelligence systems and external contract workers.
The ver.di trade union, which represents the employees, has been trying to engage with TikTok management for weeks. Kalle Kunkel, a union spokesperson, explained that they presented a list of demands for fair severance and an extended notice period of one year for the affected staff. TikTok's response has been a flat-out refusal to negotiate.
“Basically, they said: ‘We don’t want to talk with you,’ so after that, we went on two strikes,” Kunkel said. “But they still don’t react to us.”
The layoffs represent a significant blow to TikTok's German operations, reducing its 400-person workforce in the capital by nearly 40%. In a statement, TikTok spokesperson Anna Sopel said the restructuring aims to “streamline workflows and improve efficiency,” asserting the company remains committed to platform safety.
A Global Shift Towards AI Moderation
The situation in Germany is not an isolated incident but rather a reflection of TikTok's global strategy. Over the last year, the company has consistently reduced its human moderation staff worldwide, placing a greater reliance on automated systems.
This trend includes several high-profile cuts:
- In the Netherlands, the company fired its entire team of 300 content moderators.
- In Malaysia, TikTok announced plans to replace around 500 moderation employees with AI.
- Reports from this past February indicated significant layoffs across its trust and safety teams in Asia, Europe, the Middle East, and Africa.
This global downsizing comes after TikTok CEO Shou Zi Chew testified before the US Congress in 2024, pledging to increase spending on trust and safety to over $2 billion and grow the team to 40,000 people. While the company claims it is investing another $2 billion this year, it has not provided updated figures on its current trust and safety headcount.
Other social media giants, including Meta and X, have also made deep cuts to their own safety teams, with Meta reportedly planning to replace 90% of its product review staff with AI.
The Human Cost and Risks of AI Moderation
Replacing human moderators, who review up to 1,000 videos a day, with automated systems is fraught with problems. Aliya Bhatia, a policy analyst at the Center for Democracy and Technology, warns that this shift will inevitably “lead to more mistakes and more harmful experiences” for users.
Union members have shared examples of AI's shortcomings, noting that TikTok's automated system has incorrectly flagged harmless content, like a rainbow Pride flag, while failing to detect actual policy violations. “AI is not able to really identify problematic pictures or videos, especially when it comes to sophisticated content,” Kunkel stated.
This is particularly risky in the European Union, which has strict regulations under the Digital Services Act. The act requires large platforms like TikTok to vigorously police their sites for harmful content or face substantial fines.
Beyond the technology's limitations, there are also concerns for the outsourced workers who will now handle some of this work. Kunkel expressed worries that these contractors may not have access to the same mental health resources and support programs that TikTok provided for its in-house staff, who are regularly exposed to graphic and disturbing content.
Union Pushes Back Against TikTok's Stance
To force TikTok to the negotiating table, the ver.di union organized two day-long strikes and protests in late July. Kathlen Eggerling, the union's lead negotiator, emphasized that the actions were necessary to demonstrate the value of the employees.
After the first strike, TikTok sent a message to its employees warning them that participating in protests during work hours required prior notification and that it was reviewing the situation with its legal team. The union quickly condemned this move.
“It seems TikTok may need content moderators to factcheck its internal communications as well. We call on management to stop intimidating strikers,” Eggerling said, highlighting that under German law, unions are not required to disclose which workers will participate in a legal strike. “Instead of spreading misinformation, TikTok should finally come to the negotiating table.”
The union has made it clear that employees will continue to rally and that a longer-term strike is a real possibility if the company continues to refuse to engage.
