IGF 2019 WS #340
In Your Face: surveillance in the age of facial recognition

Relevance to Theme: The global diffusion of surveillance apparatuses and devices walks hand in hand with the fast paced development of new technologies for personal identification. Facial recognition is one such technology. Often presented as a solution for strengthening security and combatting fraud, as well as a useful tool for personalized publicity, facial recognition is said to offer a wide range of applications. None of the justifications for their use, however, considers how these technologies actually work, nor the complications that their implementation by market actors and public authorities brings to the fore.

Applications of facial recognition technologies at the Global South already include their use in transportation, schools, airports, stores and malls, street level surveillance, among others. The common factor among all these uses is the attempt to achieve greater social control through constant monitoring. Despite taking privacy intrusions to another level, these implementations are usually not followed by any type of regulation or contract that specifies how data collection is made and how data will be used. Consent is another absent element when collection occurs at the environmental level.

In addition to being subject to deviations on their goals, illegitimate trade of personal data, security breaches and data leaks, recent studies emphasize the race and gender biases found on facial recognition algorithms and databases, which potentially lead to the discrimination and social exclusion of already marginalized groups. This is particularly acute in the contexts of Global South countries, which are marked by stark inequalities. For instance, the faces of black women are more likely to not being recognized by facial recognition algorithms, thus leading to a higher probability of generating "false positives" in fraud checks.

The workshop aims at bringing together specialists to discuss the use of facial recognition technologies trying to build a narrative on their origins, uses, impact and actors involved in their production and distribution as well as their interests. Considering that there is a lack of literature on this subject, the workshop will focus on the Global South -- from a Global South perspective -- and on the particularities of the implementation of facial recognition in these countries and populations.

Relevance to Internet Governance: Internet governance has recently turned to the problem of massive data collection and its impacts. Mass surveillance and data breaches scandals have proliferated in the past five years, often involving public and private actors alike. This has shifted the attention to the roles of corporations and states in developing new surveillance technologies, as well as to the partnerships among them. Of particular concern is the fact that although state surveillance is not new (nor necessarily abusive), some countries have historically seen its use against marginalized populations and political opposition. Latin America, in particular, has a tradition of policies of persecution of "internal enemies", especially during the military dictatorships many countries have faced. Surveillance and repression targeted indigenous and black populations, students, academics, activists, among others. Democratic periods have also been marked by abuses in in countries like Brazil, Guatemala and Paraguay.

Public-private surveillance partnerships have grown in scope and achieved new forms, particularly as many states lack the capacity to store and process massive volumes of digital data collected on a daily basis. As new potential uses of surveillance technologies emerge, new justifications for their use are created. The strength of such discourses lays on centuries of an idea of development and neutrality of science and technology that perpasses global imaginary and -- among other factors -- allows large acceptance without questioning.

Facial recognition technologies are ultimately based on the collection of unique personal information and raise the stakes around discussions about data governance, since they depend on environmental collection of sensitive data. This brings new challenges to the debates around the idea of informed consent and other elements of the privacy self-management model. When implemented in contexts characterized by deep structural inequalities, they might pose serious risks to democracy, particularly when interacting with local state discriminatory practices. In its account of data governance and security, Internet governance debates should consider the specificities of each case, as they might significantly shape how Internet-enabled technologies are used in the Global South.