A/HRC/38/35 28. The scale and complexity of addressing hateful expression presents long-term challenges and may lead companies to restrict such expression even if it is not clearly linked to adverse outcomes (as hateful advocacy is connected to incitement in article 20 of the International Covenant on Civil and Political Rights). Companies should articulate the bases for such restrictions, however, and demonstrate the necessity and proportionality of any content actions (such as removals or account suspensions). Meaningful and consistent transparency about enforcement of hate speech policies, through substantial reporting of specific cases, may also provide a level of insight that even the most detailed explanations cannot offer.80 29. Context. Companies emphasize the importance of context when assessing the applicability of general restrictions.81 Nonetheless, attention to context has not prevented removals of depictions of nudity with historical, cultural or educational value; 82 historical and documentary accounts of conflict;83 evidence of war crimes;84 counter speech against hate groups;85 or efforts to challenge or reclaim racist, homophobic or xenophobic language.86 Meaningful examination of context may be thwarted by time and resource constraints on human moderators, overdependence on automation or insufficient understanding of linguistic and cultural nuance.87 Companies have urged users to supplement controversial content with contextual details, but the feasibility and effectiveness of this guidance are unclear.88 30. Real-name requirements. In order to deal with online abuse, some companies have “authentic identity” requirements;89 others approach identity questions more flexibly. 90 The effectiveness of real-name requirements as safeguards against online abuse is questionable.91 Indeed, strict insistence on real names has unmasked bloggers and activists using pseudonyms to protect themselves, exposing them to grave physical danger. 92 It has also blocked the accounts of lesbian, gay, bisexual, transgender and queer users and activists, drag performers and users with non-English or unconventional names. 93 Since online anonymity is often necessary for the physical safety of vulnerable users, human rights principles default to the protection of anonymity, subject only to limitations that would protect their identities.94 Narrowly crafted impersonation rules that limit the ability of users to portray another person in a confusing or deceptive manner may be a more proportionate means of protecting the identity, rights and reputations of other users.95 31. Disinformation. Disinformation and propaganda challenge access to information and the overall public trust in media and government institutions. The companies face 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 Julia Angwin and Hannes Grasseger, “Facebook’s secret censorship rules protect white men from hate speech but not black children”, ProPublica, 28 June 2017. See paras. 52 and 62 below. Twitter, “Our approach to policy development and enforcement philosophy”; YouTube policies (the importance of context); Richard Allan, “Hard questions: who should decide what is hate speech in an online global community?”, Facebook Newsroom, 27 June 2017. Submissions by OBSERVACOM, p. 11 and ARTICLE 19, p. 6. WITNESS submission, pp. 6−7. Ibid. Electronic Frontier Foundation submission, p. 5. Association for Progressive Communications submission, p. 14. See Allan, “Hard questions”. YouTube policies (the importance of context); Facebook community standards (hate speech). Facebook community standards (using your authentic identity). Note that Facebook now permits exceptions to its real-name policy on a case-by-case basis, but this has been criticized as insufficient: Access Now submission, p. 12. Baidu even requires the use of personally identifying information: Baidu user agreement. Twitter Help Center, “Help with username registration”; Instagram, “Getting started on Instagram”. J. Nathan Matias, “The real name fallacy”, Coral Project, 3 January 2017. Access Now submission, p. 11. Dia Kayyali, “Facebook’s name policy strikes again, this time at Native Americans”, Electronic Frontier Foundation, 13 February 2015. See A/HRC/29/32, para. 9. Twitter rules and policies (impersonation policy). 11