TikTok, one of the popular short-format video sharing platforms from China, which gained popularity worldwide (but is banned in India now along with other Chinese apps), has informed employees based in the UK- London trust and safety division, about the potential job losses, according to a Financial Times report. The email was sent to them on Friday (August 22) stating that moderation and quality assurance work may no longer be handled in London. The decision comes shortly after the Online Safety Act took effect, raising concerns about reduced human involvement in monitoring harmful and illegal content.
The company currently employs over 2,500 people in the UK, but “several hundred” roles are likely to be cut both in the UK and across South and Southeast Asia. A town-hall meeting was held with affected employees to discuss the company’s next steps.
Online safety act and its impact
The timing of the layoffs comes just weeks after the Online Safety Act 2023 came into force. The law requires platforms like TikTok, Facebook, YouTube and X to protect children online, remove illegal content and regulate harmful material.
However, the Act has been controversial. Over 468,000 people signed a petition seeking its repeal, claiming the rules encourage over-censorship and restrict free speech. Critics have also opposed strict age verification checks, which require users to upload personal data to access adult content. Despite objections, the UK government and regulator Ofcom are pushing ahead with the rollout.
TikTok’s AI-driven moderation push
TikTok says the move is part of a global reorganisation to strengthen efficiency through AI. The email to staff cited “technological advances, such as the enhancement of large language models” as a reason for shifting moderation work away from London.
The platform is also introducing machine learning tools for “age assurance,” inferring user age from online activity patterns. However, regulators have yet to approve these tools.
Experts warn that over-reliance on AI moderation could pose risks, including reduced accuracy, cultural insensitivity, and failure to detect harmful content in local contexts. Human moderators, critics argue, remain crucial for ensuring safety, fairness, and context-aware content evaluation.