Similar worrying gaps in Meta’s enforcement in India are also prevalent in content promoting bullying, harassment, violence, and incitement. India has Meta’s largest user base in the world.
The degree of Meta’s enforcement on different kinds of posts or accounts which contain content violating community standards – ranging from hate speech, drugs, terror, nudity, sexual activity, child endangerment, and more – is measured through the reported ‘proactive rate’, which is disclosed in its transparency reports. The time between a piece of content being flagged by users to when Meta takes action on the post is varied and not revealed in the reports. This, rather, points to Meta’s reactive moderation relying on user flagging rather than it’s proactive automated scanning and removal of harmful content falling under the above-listed three groups; which is not the case with other forms of harmful content that it proactively moderates (like nudity, child endangerment, terrorism), as shown in the graphic above. Thus, in several cases, such detrimental posts can remain on the platform for an extended period of time.
For example, a proactive rate of 30% for hate speech means that out of all the posts Meta took action on related to hate speech, only 30% of those posts were flagged by Meta before they were reported by users.
Meta describes this ‘proactive rate’ as “the percentage of all content or accounts acted on” before users report them to Meta.
