In an attempt to justify and backup the Mark Zuckerberg’s claims of Facebook’s intense efforts to kibosh objectionable content, the tech giant released a Facebook Community Standards Enforcement Report.
The maiden report spans across 81 pages and talks in detail about the metrics on enforcement on six different key areas namely graphic violence, adult nudity, and sexual activity, terrorist propaganda (ISIS, al-Queda and affiliates), spam, hate speech and fake accounts.
The data contained in the report analyzes the period between October 2017 and March 2018.
Enlisted below are some of the major statistics highlighted in the report:
- As per the report, Facebook has deleted 837 million pieces of spam content received in the first quarter of 2018, and almost 100% of the total posts were flagged and removed by Facebook systems before anyone flagged it.
- As many as 583 million fake accounts were disabled in the Q1 2018. The report mentions that most of the fake accounts were taken down within the minutes of their creation. Besides, estimated 3 to 4% of the total active accounts were still fake.
- 3.5 million pieces of content that violated the guidelines of Facebook and were termed as violent content, were removed from the portal. 86% of the total taken down posts were disposed of by the intuitive Artificial Intelligence system before any user flagged them.
- 21 million posts featuring adult nudity and sexual activity were axed int he first quarter of 2018, and out of them 96% posts were automatically detected by the smart AI system of Facebook.
- Facebook caught and removed 1.9 million pieces of content that were proliferating terrorist propaganda in the opening quarter of 2018. Interestingly, this figure was 72% more than the figure recorded during the previous year.
- Mark Zuckerberg was questioned heavily for the counter mechanism to stop hate posts in his much talked about appearance before the officials. The report explicitly highlights the Facebook’s inability to identify and taken down posts containing any form of hate speech with the help of automated systems. As per the report, 2.5 million pieces suspected for hate speech were removed in Q1 2018, and their AI system reported just 38% of them.
- A surge of 4% has been witnessed in the requests received by the officials of different countries for account data. Facebook received 82,341 requests to disclose the account information of users. The US topped the list with 53625 requests and India stands on the second position with 17262 requests.
The report is a clear indication of the fact that Facebook is committed to making the platform an amiable place for users. Although the fact that report spurts just numbers and no examples were displayed to back the numbers. This jeopardizes the credibility of the report.
Moreover, through this report, Facebook has clearly self-praised its AI system and has reassured the Zuckerberg’s statements in his hearing before the US Senate regarding Artificial Intelligence being one of the most efficient methods to filter the large number of contents.
Have something to add? Share your views and keep reading Fossbytes.