國際人權公約對網路中介的適用：以性別仇恨言論為中心

Abstract

性別仇恨言論是煽動基於性別的歧視、敵視或暴力的言論，為公民與政治權利國際公約所不許，締約國有義務以法律禁止之。然而，當今日益兩極化的輿論中，對是否應禁止仇恨言論仍無共識。網路科技的發展，則無疑加劇仇恨言論傳播速度與國際影響力。雖然社群網路公司已經採取行動加以審查，卻也導致一些使用者遷移至承諾不審查的網站，使其可以不受限制的散佈仇恨。個別案例中，甚至對人身安全導致立即性的危險。 本文無意處理社群網路上的言論審查問題，而專注於一項未受廣泛矚目的議題，即提供網路基礎建設服務的網路中介，包括內容遞送網路、雲端託管服務、域名註冊服務等，是否基於國際人權公約而負有義務，應建立符合國際人權標準的自律規範，以減緩性別仇恨言論對女性與性少數的負面影響。 首先，本文網路技術出發，以技術階層的概念，系統化說明不同網路技術間的上下依賴關係。即便網站拒絕刪除違反公約的性別仇恨言論，提供網路基礎建設的網路中介仍可以透過技術手段干預之，且與國家級網路審查機制無涉。 其次，本文論證網路中介適用國際人權公約的法律基礎與劃定適用範圍，包括表意自由、禁止仇恨言論、隱私權、消除對婦女一切形式歧視公約，並以案例說明應如何解釋適用上開公約保障之權利。本文亦比較美國、德國和歐盟規範網路中介的內國法與國際人權標準的差異，作為落實本文主張之參考。 最後，本文主張網路中介技術上有干預能力，規範有義務直接適用國際人權公約處理性別仇恨言論，建立符合國際人權標準的自律規範。

Gendered hate speech is any advocacy that incites discrimination, hostility, or violence on the basis of gender. Such speech is prohibited under the International Covenant on Civil and Political Rights (ICCPR), and the States Parties are obliged to prohibit it by law. However, in today's increasingly polarized society, there is no common consensus on whether and how hate speech should be regulated. With the rapid development of the Internet, technology has undoubtedly increased the spread and international reach of hate speech. While social media companies have taken actions to bar hate speech, their efforts have led some users to migrate to other sites that promise to uphold so-called "free speech", allowing them to spread hate without restriction and, in some cases, pose an immediate danger to personal safety. This thesis does not attempt to address the issue of censorship on social media. Instead, it focuses on a nuanced issue that has yet to receive much attention, namely, whether Internet intermediaries that provide Internet infrastructure services, such as content delivery networks, hosting providers, and domain name registrars, are obligated under international human rights law to establish a self-regulation framework that complies with international human rights standards in order to mitigate the impact of gendered hate speech towards women and sexual minority. First, this thesis begins with technical analysis and employs the concept of tech stack to systematize the dependency relationship between different network technologies. Even if a website refuses to remove gendered hate speech that violates ICCPR, Internet intermediaries that provide the network infrastructure can still intervene through technological means. Such intervention requires no state-sponsored internet censorship mechanisms. Second, this thesis establishes the legal basis for applying international human rights law to private Internet intermediaries and sets the scope of application, including freedom of expression, prohibition of hate speech, the right to privacy, and the Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW). Followed by an application on real-life cases to demonstrate how the rights protected by the international human rights law should be interpreted and applied. It also compares the differences between international human rights standards and national laws regulating Internet intermediaries in the United States, Germany, and the European Union as a reference for implementing the propositions made in this thesis. Finally, this thesis concludes that Internet intermediaries are technically capable of intervening and are obligated to establish a framework for self-regulation that complies with international human rights standards and to apply international human rights law directly to address gendered hate speech.