Instagram Fights Mean Comments with AI: A New Era for Online Safety?
Instagram’s new AI system is built upon the success of their existing spam filter, which utilizes DeepText to identify and block unwanted content. This system has been trained with human input on identifying negative comments and categorizing them into groups such as bullying, racism, or sexual harassment. The AI system has been rigorously tested with over two million comments analyzed by human raters, ensuring its accuracy.
When a comment is flagged by the AI, it is automatically removed from the user’s feed, although the original poster can still see their comment. This prevents users from circumventing the filters through repeated attempts. The filter is currently available in English only, but Instagram is expanding its spam filter to other languages, suggesting a broader rollout of this AI system in the future.
This move by Instagram reflects their commitment to creating a safer online community. While the internet often serves as a platform for hurtful language, Instagram hopes this AI system will help create a more positive and inclusive space for its users.
