In the third quarter, technology improvements allowed the publication of more than 22 million such posts to be blocked in advance.
Facebook Inc, which owns the social network Facebook, on Thursday for the first time released statistics demonstrating the prevalence of hate speech on its platform. According to the company, in the third quarter of 2020, 10,000 views of content published by users accounted for 10-11 messages containing hate speech.
The world’s largest social network, now under the scrutiny of the authorities, included an assessment of the share of publications containing xenophobic and discriminatory statements in its quarterly report on content moderation.
In the third quarter, Facebook moderators took measures against 22.1 million posts whose authors incited hatred, and the vast majority of publications were removed in advance, that is, even before receiving complaints from disgruntled users. In the previous quarter, 22.5 million such posts were moderated. By “taking action” on Facebook we mean removal of posts, warnings to users, disconnection of accounts or transfer of information about such publications to third parties.
Another popular social network Instagram, owned by Facebook Inc., moderated 6.5 million publications, the authors of which tried to incite hatred. In the second quarter, for comparison, 3.2 million such posts were removed from Instagram. About 95% of them were identified in advance, which is 10% more than in the previous quarter.
Last summer, civil rights groups organized a large-scale boycott by advertisers against Facebook, trying to put pressure on the social network and force its leadership to strengthen the fight against hate speech.
In October, the social network’s leadership reported a reform of its policy on hate speech publications. In particular, it was about removing all posts whose authors deny or distort the essence of the Holocaust. Thus, the company made a 180-degree U-turn after comments previously made by Facebook CEO Mark Zuckerberg about what is allowed to be published on the social network.
Facebook deleted 19.2 million posts containing scenes of violence in the third quarter, compared to 15 million deleted posts in the previous quarter. Instagram deleted 4.1 million such publications, compared to 3.1 million posts moderated between April and June.
Earlier this week, Congressmen criticized Zuckerberg and Jack Dorsey, CEO of Twitter Inc, for the policy of moderating posts published in social networks. Among the accusations against social networks were statements by Republicans about political bias on the part of technological giants against conservatives.
Recently Facebook has also been criticized for allowing the rapid growth of groups whose participants spread misinformation about elections and aggressive rhetoric.
The report says that the rate of detection of banned content has increased due to the improvement of artificial intelligence tools that allow you to find and remove posts with hate speech even before receiving complaints from users outraged by certain publications.
According to data published in the report, from July to September the social network also removed 12.4 million posts containing child pornography. This figure also increased compared to the previous quarter, when moderators took action on 9.5 million such publications.