por Eduarda Costa e Felipe Silva
On 11 November, 2020, during the Internet Governance Forum – IGF 2020’s Community-facilitated Networking Break, LAPIN promoted a multidisciplinary debate about the impacts of Facial Recognition Systems on privacy and data protection under the title “Surveilling the surveillants: monitoring facial recognition deployment in Latin America and Caribbean”.
The discussion was based on three main questions previously elaborated by LAPIN. Those questions were:
(i) where and for what purposes is facial recognition being used in your region (shops, banks, public security, marketing, government services, building access)? Are there any specific cases you would consider remarkable to share with us?
(ii) What are the initiatives about facial recognition that your organisation or other organisations from your country/region are currently working on? What are the main tactics your organisation applies to obtain information about the technology from those who are deploying it?
(iii) Is there any bill or legislation in your region/country for regulating the use of facial recognition?
With regard to the first question, participants highlighted cases in which facial recognition systems have been deployed in metro and bus stations, for both security and advertising purposes, as well as for predictive policing. The use of the technology in metro stations in São Paulo and Salvador, as well as tests carried out during the carnivals of Rio de Janeiro and Brasilia, were among the cases discussed.
Participants expressed concerns involving the accuracy of facial recognition technology (FRT). They mentioned cases in which citizens were misidentified and mistakenly conducted to police stations by officers, who only realised that there had been a misunderstanding afterwards. Such a case happened in Rio de Janeiro, for instance, in 2019. Such a case happened in Rio de Janeiro, for instance, in 2019.
Regarding the second question, the opacity of facial recognition systems and the lack of information provided by public bodies were common issues faced by the participants when trying to obtain data about such use cases. Despite that, participants mentioned the use of semi-structured interviews and official requests made under the Brazilian Information Access Legislation (LAI) addressing public authorities as particularly useful tools for obtaining more information on how the state was deploying this technology.
Furthermore, the European civil society coalition Reclaim Your Face was also highlighted as an interesting initiative to protest against the spread of FRTs adoption.
Finally, after identifying an increasing deployment of facial recognition by both government and the private sector, participants indicated that some measures should be taken regarding the high risks that this technology represents to society. The main tactics for addressing FRT risks were: (i) the necessity of specific legislation, (ii) adopting regulation that aims to minimise the bias of these technologies regarding race and gender, and (iii) a moratorium, since the impact and harm of facial recognition cannot be completely foreseen.
