[ad_1]
Currently, most deepfakes found online are of NCII, which disproportionately feature women. As Gregory pointed out, much of the concern surrounding deepfakes is centered on how they can be weaponized against political figures in the electoral context, but many seem to forget that deepfakes are already being used to victimize private individuals. In a workshop convened by Witness in São Paulo, Brazil, the participants underscored that deepfakes perpetuate gender-based violence and can be used, among other things, to degrade, threaten, humiliate, or bully women. For instance, a deepfake sex video of a journalist in India was made in order to silence her after she had called for immediate action to be taken against child sex abusers following the gang-rape of an eight-year-old Kashmiri girl.
Gregory spoke of threat prioritization, the need to invest our time and energy into tailoring solutions that address the existing and emerging forms of harm wrought by deepfakes and other types of synthetic media, as part of Witness’s 12-step approach to prepare for deepfakes. Step 3 of Witness’s 12-step approach requires us to shape our responses in accordance with human rights. Given that, what are the relevant international human rights law standards implicated by NCII?
1. Freedom of expression
The right to freedom of expression is enshrined in Article 19 of the International Covenant on Civil and Political Rights (ICCPR). Under Article 19(3), restrictions on this right are permissible so long as they satisfy the (i) legality, (ii) legitimacy, and (iii) necessity and proportionality requirements. Gregory explained that prong (i) calls for clarity—that a piece of legislation enacted by a State or a content moderation policy adopted by a social media company clearly stipulates the types of speech that are prohibited. Meanwhile, prong (ii) necessitates that there be an acceptable goal or rationale behind the legislation or content moderation policy, and prong (iii) requires that there be no alternative, less intrusive measures available to achieve that goal.
A 2017 joint statement from the UN Special Rapporteurs on freedom of expression and on violence against women emphasized that online gender-based abuse and violence can force women to withdraw from online forums, negatively impacting their right to freedom of expression. And in a recent report, the UN Special Rapporteur on freedom of expression, Irene Khan, similarly warned that “[o]nline gendered disinformation campaigns are increasingly being used to deter women from participating in the public sphere….” Although vague laws and policies about misinformation or nudity would probably not satisfy the tripartite test of Article 19(3), Gregory noted that, as far as social media companies are concerned, many already have adequate policies in place to tackle the issue of deepfakes in a human rights-compliant way. The problem, however, lies in their inconsistent and slipshod enforcement, which potentially discriminates against women.
2. Right to privacy
Aside from free speech concerns, there may also be concerns for the right to privacy. Article 18 of the ICCPR prohibits any “arbitrary or unlawful interference with [a person’s] privacy, family, home or correspondence….” In a resolution adopted in 2016, the UN General Assembly recommended that States further develop or maintain preventive measures and remedies for violations of the right to privacy, particularly for women and children.
Some of the solutions for synthetic media have impacts on content creators’ right to privacy. Witness has been at the forefront of advocating for authenticity infrastructure, discussed further below, as a solution for addressing synthetic media but recognizes that such a solution may compromise privacy. Protecting women’s right to privacy should not necessarily have to come at the expense of risking others’ right to privacy, so the question of whose privacy should ultimately be protected is a fine one that we will need to tread carefully.
3. Women’s rights
Finally, because most deepfakes of NCII disproportionately impact women, it is paramount that women’s rights under the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW) are safeguarded. Article 2 of CEDAW requires States to take all appropriate means to eliminate discrimination against women. Not only must States refrain from discriminating against women, but according to General Recommendation No. 35, they must also perform their due diligence in preventing, investigating, and punishing acts of gender-based violence perpetrated by private actors, including acts that occur on the Internet. Without accountability, the dangers posed by online NCII persist.
Any solution offered for tackling deepfakes and other forms of synthetic media will need to bear these human rights principles in mind. What, then, does Gregory propose as viable solutions to this problem?
Solutions
[ad_2]
