Highlights:
1. YouTube has removed 1.9 million videos in India for violating community guidelines between January and March 2023.
2. More than 8.7 million channels were removed by YouTube in the same period for violating spam policies, including scams, misleading metadata, and spam comments.
3. YouTube also deleted more than 853 million comments, with the majority being spam.
4. The company has invested heavily in its policies and products to ensure the safety of the YouTube community.
5. Machine-learning technology was used to identify and remove the majority of the videos, with 69% of the identified videos seen less than 10 times.
YouTube, the popular video sharing platform, has removed 1.9 million videos in India for violating community guidelines. These videos were deleted between January and March 2023 under the supervision of Google, which owns YouTube. At an international level, more than 6.48 million videos have been removed for policy violations by major video streaming platforms.
According to reports, YouTube has also removed over 8.7 million channels between January and March 2023. These channels were removed for violating YouTube’s spam policy, which includes scams, misleading metadata, and spam content in thumbnails, videos, and comments. YouTube has removed over 853 million comments, most of which were spam. More than 99% of these comments were identified and removed automatically.
In a statement, YouTube stated that it has invested heavily in necessary policies and products to ensure the safety of its community. The majority of creators now upload content with confidence, following the policies, and YouTube believes that its educational efforts have been successful in reducing the number of creators who unintentionally violate the policies.
The company also revealed that over 93% of the videos removed were identified by machines rather than humans. Among these machine-identified videos, 38% were removed even before they were viewed once, and 31% were viewed 10 or fewer times before removal. This indicates that machines successfully identified more than 69% of these videos that were viewed less than 10 times.