TikTok Starts Automating Content Moderation On The Social Network


TikTok representative told The Verge that, seeking to streamline content moderation processes and reduce collaborators’ contact with harmful material, the social network will begin to adopt automated classification systems in the United States and Canada. The implementations, he explains, will take down videos related to nudity, sex, violence, illegal activities and other topics.

With the measure, he adds, professionals will be able to better dedicate themselves to the analysis of misinformation and requests from users who disagree with the removal of posts from the platform, as well as being more protected against consequences similar to those of post-traumatic stress, already identified in teams from other countries. companies.

Finally, one of TikTok’s goals is also to provide more transparency to your decisions. However, some challenges face the giant’s movement, which, in its history, has already faced multiple accusations of discrimination.

Successes and revisions

Hundreds of thousands of videos can “fall” by mistake, since in the first quarter of 2021 alone, more than 8.5 million of them went off the air just in the country run by Joe Biden – and the automated tool has a hit rate of 95%, according to TikTok. In other words, 5% would be at the mercy of unfair sanctions.

In any case, the company defends that it will apply the method only in what is most reliable. Furthermore, he claims that humans will determine the fate of most disputed cases. In the coming weeks, when the news will be distributed, we will know if the change is really effective.


Please enter your comment!
Please enter your name here