According to a presentation prepared by staffers at Facebook—which has since been re-branded Meta—there were several spikes in inflammatory content in India between November 2019 and July 2020. A two-part internal presentation titled “Communal Conflict in India,” dated 14 July 2020, noted a marked increase in such content in English, Hindi and Bengali in the backdrop of the protests against the Citizenship (Amendment) Act and the nationwide lockdown imposed in the first few months of the COVID-19 pandemic. The insights reflected in the presentation add to mounting questions around Facebook’s responsibility for the proliferation of communal content in India and the efficacy of the company’s efforts to tackle these issues within its largest market. According to the presentation, at the time that the report was collated, nearly two-thirds of those with internet access in India used apps owned by Meta. “If social media survives 10 more years like this, there will only be hatred,” a Muslim man from Mumbai who was interviewed for the presentation told the researchers. “If social media policies don’t change,” he added, “I would say it would be very difficult place to survive for everyone.”
This internal presentation is a part of the disclosures made by the legal counsel of Frances Haugen—a former product manager in Facebook’s civic-integrity team—to the United States’ Security and Exchange Commission, with which she filed a series of complaints about the technology conglomerate’s role in perpetuating misinformation, hate speech and violent extremism. Haugen’s lawyers provided legislators with redacted versions of these documents, which were reviewed by a consortium of news organisations, including The Caravan.
Facebook’s internal figures on inflammatory content in India, as well as user-reported hate content, showed surges through late 2019 and early 2020. According to the presentation, the company recorded that “inflammatory prevalence” in India shot up several times since November 2019—particularly for Hindi and Bengali content in late March 2020—with a rise of three hundred percent over the baseline for such content. The presentation stated that based on figures from June 2020 for Hindi and Urdu, these levels were comparable to Pakistan but higher than most other countries that Facebook denotes as at-risk for conflict or violence. The baseline for such content on the platform was unclear and a Meta spokesperson did not respond to questions relating to it.
Similarly, user-reported hate content spiked in December 2019, March 2020 and May 2020—with increases of over eighty percent compared to the baseline. Between June 2019 and July 2020, there was a significant increase in action rate on user-reported hate content. This happened especially for content in English, which was being actioned at a rate four times higher than that on 1 June 2019. According to the presentation, this increase could be attributed to an increase in the rate of action being taken against content that was being reported rather than more reports being filed.