Meta response to the Business & Human Rights Resource Centre, August 31, 2022

"We’re aware it’s vital our services work to detect potential hate speech, and to prevent it from spreading. We’ve invested significantly in technology to find hate speech in various languages, including several regional languages, such as Hindi and Bengali. As a result, we’ve reduced the amount of hate speech that people see by half in 2022. As of the time of writing, it’s down to 0.02 percent on Facebook. Sadly, hate speech against marginalized groups is on the rise globally. So we are improving enforcement, and we are committed to updating our policies as hate speech evolves online.

We have reviewers across 20 languages commonly spoken in India—accounting for 83% of the population—with both language and cultural expertise. Our human reviewers review content in the following languages: Hindi, Bengali, Tamil, Malayalam, Punjabi, Urdu, Kannada, Marathi, Gujarati, Assamese, Telugu, Oriya, Sindhi, Mizo, Marwari, Chhattisgarhi, Tulu, Maithili/Bhojpuri, Konkani and Meitei. In addition, we have invested in setting up a robust third-party fact-checking network in India. Meta has 11 fact-checking partners in India (the highest country coverage in the world), and these fact-checkers have the ability to rate content in 15 languages.

You and other stakeholders may be interested to note we already issue a monthly India enforcement report, in accordance with Indian law, which captures our enforcement data for both proactive actions and user complaints; those reports are publicly available here. We hope stakeholders and rights-holders will find them useful as they monitor relevant content and enforcement trends.

It’s also important to note that Meta released insights and actions from its India human rights impact assessment as part of its annual Human Rights Report (pp. 57-60). The disclosure seeks to follow best practice guidelines as stipulated by the UN Guiding Principles on Business and Human Rights and the reporting guidance from the Danish Institute for Human Rights Guidance on HRIA of Digital Activities. The disclosure includes sections on insights, risks, and actions. It shares rich detail about salient human rights risks — as well as about steps Meta is taking to prevent or mitigate them. We hope all stakeholders and rights-holders will find these sections useful.

Finally, the document you shared with us makes recommendations regarding our Dangerous Organizations and Individuals policies. While we cannot discuss specific cases, our policies warn that Meta may ban violating individuals or entities after following a careful, rigorous, and multi-disciplinary process. We enforce our Dangerous Organizations and Individuals policy globally. You can access these policies online here; and they are, like other content policies, continually updated."