Whistleblower alleges RSS promotes fear, anti-Muslim narratives on Facebook
Facebook whistleblower Frances Haugen has kicked up a storm in the US with her leaks highlighting how Facebook's internal decisions are detrimental to individual well-being and society at large. Fresh reports indicate her complaints to the US Securities and Exchange Commission (SEC) reference alleged fear-mongering and dehumanizing content promoted by Facebook accounts purportedly run by the Rashtriya Swayamsevak Sangh (RSS). Here are more details.
Haugen leaked vast amounts of confidential Facebook research, data
Haugen submitted around eight complaints to the SEC to prompt regulation of her former employer, Facebook. She also leaked thousands of pages of internal research and documentation to the Wall Street Journal for The Facebook Files. Haugen's revelations forced Facebook to halt the development of Instagram Kids, perhaps shaved around $6 billion off CEO Mark Zuckerberg's value, and triggered Congressional and Senate Committee hearings.
What exactly are the allegations against RSS?
According to Hindustan Times, the complaint filed with the SEC reads, "RSS (Indian nationalist organization Rashtriya Swayamsevak Sangh) Users, Groups, and Pages promote fear-mongering, anti-Muslim narratives targeted pro-Hindu populations with V&I (violence and inciting) intent." The RSS is widely regarded as the originator of ideologies for the ruling Bharatiya Janata Party (BJP).
India one of Facebook's 'top three political priorities' in 2020
The document titled Adversarial Harmful Networks—India Case Study submitted alongside the complaint cites "political inconsideration" and the categorization of India in "Tier zero" among Facebook's "top three political priorities". The document also mentions Facebook's damningly acute awareness about content against a particular religion in India. According to internal assessments, only 0.2% of reported hate speech is taken down.
Facebook maintains it actively combats hate speech in Indian dialects
Meanwhile, Facebook claimed that it now detects and eliminates more hate speech than ever before. In 2019, it claimed that it supported four Indian dialects—Hindi Bengali, Urdu, and Tamil. However, Haugen's complaint alleges that Facebook doesn't have classifiers in place that would flag ill-intentioned content being propagated in Bengali. Facebook has since faced allegations of inaction against content posted by certain groups in India.
Internal records suggest Facebook cannot flag Hindi, Bengali hate speech
The complaint claims that Facebook's internal records show how the lack of Hindi and Bengali classifiers (algorithms that detect hate speech) meant that Hindi and Bengali content pushing the anti-Muslim narrative was never flagged by systems and dealt with. The significance of this failure is amplified by the fact that 340 million of Facebook's 2.89 billion global users are in India.
Facebook classifies India in 'Tier zero' alongside US for elections
As for the aforementioned "Tier 0" electoral classification for India alongside US and Brazil, the complaint doesn't explain what it means. If it helps, the subsequent "Tier 1" includes countries such as Germany, Indonesia, Israel, Iran, and Italy. Facebook's inaction on removing certain accounts and mishandling account duplicity is also under the scanner in this complaint Haugen filed with the SEC.
Complaint notes Facebook didn't act on complaint against Bajrang Dal
For context, last year a WSJ report cited an internal report and termed the Hindu nationalist group, Bajrang Dal, as a "dangerous" organization. Facebook reportedly remained indecisive on the report because, according to Facebook's security team, action against Bajrang Dal could "endanger" their business prospects and staff in India. WSJ subsequently reported that Facebook "balked at removing the group."
Allegations strengthen case against Facebook allowing amplification of political messaging
Regarding duplicate accounts, Haugen's complaint alleges Facebook is doing little to counter the problem. The data Haugen shared suggests Facebook was aware of one "Lotus Mahal," indicative of duplicate accounts being used by affiliates of the ruling BJP government. Her complaint reportedly mentioned that BJP IT Cell workers shared coordinated messaging instructions with supporters for a campaign targeting politically sensitive tags.