8.1 META’S ROLE IN HUMAN RIGHTS HARMS In seeking to assess whether a company has reached the threshold of ‘contributing’ to an adverse human rights impact, a variety of factors will be taken into account, and will vary depending on the context and the business activity. The author of the UN Guiding Principles, Professor John Ruggie, has explained: “There is a continuum between contribution and linkage. A variety of factors can determine where on that continuum a particular instance may sit [including] the extent to which a business enabled, encouraged, or motivated human rights harm by another; the extent to which it could or should have known about such harm; and the quality of any mitigating steps it has taken to address it [emphasis added].”312 The OECD Due Diligence Guidance (2018) further explains: An enterprise “contributes to” an impact if its activities, in combination with the activities of other entities cause the impact, or if the activities of the enterprise cause, facilitate or incentivise another entity to cause an adverse impact.313 The Facebook platform’s core functionality – as a social media platform enabling mass dissemination of communications by users – enabled other parties to disseminate anti-Rohingya content, including advocacy of hatred that incited violence and discrimination. This was especially true in Myanmar, where Meta was so completely dominant that its Facebook platform was most people’s primary news source and became synonymous with the internet. Meta’s business model relies on the invasive collection and analysis of vast amounts of data about people in the digital world, which can then be monetised for the purpose of targeted advertising. Meta’s core product mechanics for Facebook – its content-shaping algorithms – are optimized to ensure that users engage with content on Facebook as much as possible and spend as much time as possible on the platform. The more engaged users are, and the more time they spend on the platform, the more advertising revenue Meta earns. Research has consistently shown that the result of Meta algorithms’ optimization for maximum engagement is that the most inflammatory, divisive and harmful content is prioritized (See Section 6.2). As outlined in Sections 6.1 and 6.2, above, Meta’s use of content-shaping algorithms which were designed to maximize engagement, including its news feed and recommendations algorithms, facilitated the dissemination of advocacy of hatred against the Rohingya. By optimizing for engagement and amplifying the most inflammatory content, in a context where anti-Muslim and anti-Rohingya sentiment was rampant, Meta effectively delivered these messages on behalf of third parties to the exact audiences which were most likely to be engaged by this content – and by extension – motivated towards supporting or engaging in mass violence. As such, the evidence outlined in the previous sections clearly supports the contention that Meta, through the deployment of its content-shaping algorithms, enabled and facilitated the dissemination of antiRohingya content and advocacy of hatred. Meta also created financial incentives for third parties to produce and share harmful content through its monetisation drive which began in 2015, and which was rolled out in Myanmar in 2016. As outlined in Section 6.4, above, Meta’s ‘Instant Articles’ feature and its associated monetization drive, along with the platform’s engagement-centric algorithms, combined to create financial incentives for ‘creators’ to publish the most engaging, and therefore inflammatory, content they could. In addition to profiting from the dissemination of inflammatory anti-Rohingya content because of its impact of Meta’s engagement metrics, Meta hosted paid advertising on Facebook from the Myanmar military and prominent anti-Rohingya hate group Ma Ba Tha, which would have enabled the groups to micro-target their audiences.314 Meta would have received direct payments from these groups or their representatives in order for these ads to have been shown. Although Amnesty International does not have direct proof of these John G Ruggie, “Comments on Thun Group of Banks Discussion Paper on the Implications of UN Guiding Principles 13 & 17 In a Corporate and Investment Banking Context”, 21 February 2017, businesshumanrights.org/sites/default/files/documents/Thun%20Final.pdf, cited in BSR, “White Paper: Seven Questions to Help Determine When a Company Should Remedy Human Rights Harm under the UN Guiding Principles”, 7 January 2021, bsr.org/en/our-insights/reportview/seven-questions-to-help-determine-when-a-company-should-remedy-human-rights [emphasis added] 312 313 OECD Due Diligence Guidance (previously cited), p. 70 [emphasis added]. 314 Victoire Rio, “The Role of Social Media in Fomenting Violence: Myanmar”. THE SOCIAL ATROCITY META AND THE RIGHT TO REMEDY FOR THE ROHINGYA Amnesty International 59