'Over 30 mn takedowns, most for violent content': FB's first report under new rules

The detailed report regarding action taken for public grievances, by Facebook-owned WhatsApp for its traceability clause, is expected to get published on July 15.

From violent and graphic content, followed by adult nudity and sexual activity to suicide and self-injury content, Facebook removed nearly 30.1 million URLs, including posts, profiles and pages, according to its first monthly interim compliance report (from May 15-June 15) published by the social media giant under the new social media and intermediary rules on July 2.

Representative Image

According to the interim report published by the company on its transparency platform, Facebook’s automated and removal of such posts was pegged at over 95% for all the 10 categories. The detailed report regarding action taken for public grievances, mandated under the contentious new guidelines challenged by Facebook-owned WhatsApp for its traceability clause, is expected to get published on July 15.

The social media firm removed 2.5 million counts of violent and graphic content, at a 99.9% proactive rate, which means that the content that had to be reported by users was less than 0.1%. Proactive removal translates into the automated removal of content.

As for adult nudity and sexual activity, a point stressed under the new rules with a 24-hour takedown requirement, the company actioned 1.8 million counts at a 99.6% proactive rate. According to people familiar with the matter, the number of counts of actioned content includes pages, profiles and posts. Actioned content includes pages, posts, profiles, posts, et cetera, against which an action is taken.

“Facebook’s India specific numbers are more detailed than offered to other countries,” a person familiar with the matter said. "The list of categories is not exhaustive

According to the social media firm’s last quarter transparency report, from January 2021 to March 2021, it actioned 31.8 million URLs for adult nudity, while in India the number stood at 1.8 million for the month. In terms of graphic and violent content, the company took down 34.3 million posts, pages and profiles, with India’s number for May 15 to June 15 being 2.5 million.

The other actioned content includes spam, 25 million counts of which were removed, dangerous organizations and individuals, organized hate wherein 75,000 such URLs were taken down.

Nearly 106,000 profiles, pages and posts were actioned for being terrorist propaganda, while nearly 311,000 hate speech URLs were also taken down.

Instagram, on the other hand, removed nearly 2,031,300 URLs. Suicide and self-injury was the most actioned category with nearly 699,000 takedowns, followed by violent and graphic content at 668,000 removals and adult nudity and sexual activity at 490,000.

The report concerning removals pertaining to user complaints will be generated by July 15 by the social media company, including for its messaging service WhatsApp. The compliance officers, grievance redressal officer and nodal officer for Facebook and Instagram are the same, while WhatsApp has appointed different people.

“Over the years, we have consistently invested in technology, people and processes to further our agenda of keeping our users safe and secure online and enable them to express themselves freely on our platform. We use a combination of Artificial Intelligence, reports from our community and review by our teams to identify and review content against our policies. We’ll continue to add more information and build on these efforts towards transparency as we evolve this report,” a Facebook spokesperson said.

Facebook is the second global tech giant after Google to publish a monthly report in keeping with the new rules. The new guidelines also seek additional safeguards from significant social media intermediaries, which have 5 million users or more, to expedite their automated removal process in order to action violative content more promptly. At Facebook, the removal process is three-tier, it can involve a user complaint, automated removal, which is followed by a human review to ensure that content actioned is for the right reasons, people familiar with the matter said.

Meanwhile, Google on June 30 said it actioned nearly 60,000 pieces of content in April, mostly for copyright infringement.