More than 50 state and local legislators from 13 states mobilized in the initiative called #SaiDaMinhaCara
Starting today, June 21, more than 50 legislators from different parties are presenting bills to ban facial recognition in public spaces. The #SaiDaMinhaCara initiative demonstrates a cross-party consensus on this technology’s invasive and discriminatory nature, mainly when applied under a purported narrative of safety. With this initiative, Brazil joins other countries in trying to restrict specific and potentially harmful and discriminatory uses of facial recognition.
Many jurisdictions outside of Brazil have passed laws limiting the use of this technology over the past couple of years. San Francisco, home to the big tech companies of Silicon Valley, was the first city in the United States to approve, in 2019, a ban on the use of facial recognition by police and other public agencies. In 2020, it was the turn of Boston and Cambridge, where Harvard and MIT, university hubs of technological debate, are located. In Europe, in October 2021, the European Parliament also voted in favor of the ban after an opinion from the European Data Protection Supervisor (EDPS).
Now, 12 states and the Federal District had bills presented in Brazil, either at the federal, state or municipal level. They are: Bahia, Ceará, Espírito Santo, Distrito Federal, Minas Gerais, Pará, Pernambuco, Paraná, Rio de Janeiro, Rio Grande do Sul, Santa Catarina, Sergipe and São Paulo.
“It is always a challenge uniting parliamentarians from different political parties, legislative bodies (federal, state and municipal) and from different regions of Brazil, a country with continental dimensions and so unequal regionally. Throughout the articulation of #SaidaMinhaCara initiative, we specially sought on talking to parliamentarians who are active in state and municipal territories, guided by the concern of rooting the necessity of banning facial recognition and making this agenda more concrete to society”, said Vanessa Koetz, Coding Rights General Project Manager.
“This is an
articulation that involved several political parties. Throughout this whole process it became even more evident that the use of facial recognition technologies is at different levels in Brazilian municipalities and states. We found situations that range from City Halls that imposed the use of facial recognition without any public debate, to concrete experiences of resistance to the use of this technology from civil society and social movements.”, said Jordana Almeida, Advocacy Strategist at Coding Rights.
There are several reasons for the ban. Studies have shown that these technologies operate with harmful biases, with particularly glaring margins of error when it comes to the faces of black people, especially if they are women or transgender people. The result is that if the police use this technology, they can wrongly identify someone, and that this wrongful identification inordinately impacts women, transgender people, and people of color.
“One of the main challenges in writing a bill to ban facial recognition is to uncover the fallacy of technological solutionism for public safety. It is important to recognize the underlying risks, particularly the reinforcement of pre-existing patterns of racial discrimination in the Brazilian criminal justice system” said Bianca Kremer, Data and Feminism fellow at Coding Rights.
In Brazil, since 2019, there has been an expansion of facial recognition implementation under the guise of public security. On the second day of testing this technology in Rio de Janeiro, a woman was arrested when she was mistaken for someone already incarcerated. In Piauí, a man was transferred to the Federal District and wrongly imprisoned for three days after a facial recognition system mistakenly flagged him as a wanted person. In Salvador, a 25-year-old with special needs was approached by police officers after being mistaken for a man wanted for robbery.
“Facial recognition has already demonstrated its dangers when under the power of police officers. People are wrongly detained, violent approaches happen, lack of knowledge to operate the system are some of the consequences we have seen. Facial recognition technologies reinforce the racist character of police action and give impetus to a mistaken idea that arresting more people represents an improvement in violence and crime rates”, said Pablo Nunes, Center for Studies on Security and Citizenship (CESeC) Adjunct Coordinator.
With the fear of being accused of fueling racism and further police violence, even more so after the murder of George Floyd, big companies like IBM, Microsoft, and Amazon no longer sell this kind of technology to state authorities and for policing purposes.
“Even tech giants, who as such hold huge databases and processing power, aren’t selling or sharing their own facial recognition software for policing purposes. Why would anyone consider that this tech is good enough to be spread across our streets? Not everything that is new or technological is good. We have already had many experiences that show that certain usages of particular technologies can be racist, transphobic, exclude people and become instruments of mass surveillance. Facial recognition deployed in public spaces is one example of that”, said Joana Varon, Coding Rights Executive Directress.
But many other companies are out there, making million-dollar contracts and deals with governments by selling a flawed technology that not only exacerbates inequalities but also costs the public sector dearly. In Bahia, the government announced the expansion of the facial recognition system to more than 70 municipalities in the countryside, costing R$ 665 million. Many of these expensive surveillance tools will be distributed in cities where resources for schools, hospitals, and access to justice services are lacking. Even when a company offers to “donate” cameras, it’s very expensive: staff costs are increased to use a technology that doesn’t work; the company profits by testing its algorithm with data from our faces; and the company can build ties with the public entity that can make it easier to win future bids on the subject.
It is not just a matter of improving the flawed and racist algorithm. Even if these technologies evolve to work perfectly, they still usher in a regime of mass surveillance in public spaces.
“Mass surveillance, besides seriously damaging people’s privacy, makes even more vulnerable the population that is not served by basic public policies such as health, education and employment”, said Debora Pio, Reseacher at Medialab/UFRJ and Rede Lavits.
In a perverse reversal of the presumption of innocence, all people exercising their fundamental right to come and go and participate in public life are now treated as suspects: filmed, watched, and potentially identified without consent. Therefore, the company ViaQuatro, which has the concession of the yellow line (4) of the São Paulo Metro, was ordered to pay R$100.000 when it was proven that there was capture of users’ image for commercial purposes without their consent. Recently, the court interrupted the installation of facial recognition system on lines blue (1), green (2) and red (3).
“In this case, the São Paulo’s Justice Court suspended facial recognition systems in the São Paulo Metro based on the argument that the system could affect citizens’ fundamental rights. The problems of this technology are multiple and the strategies to ban it must also be diverse. The legislative route emerges as a battlefront for restrictions on facial recognition uses in public spaces.” said Luã Cruz, Researcher at the Brazilian Consumer Defense Institute (Idec). These are some of the concerns addressed by the #SaiDaMinhaCara initiative.
The #SaiDaMinhaCara initiative is sparkled by Coding Rights, MediaLab-UFRJ/Rede Lavits, the Brazilian Consumer Defense Institute (IDEC), and the Center for Studies on Security and Citizenship (CESeC), organizations specialized in technology, security, and human rights, which have engaged with parliamentarians around the issue. As a result, the first bills were filed on December 08, 2021, by State Representative Dani Monteiro — PSOL RJ (President of the Human Rights Commission of the Legislative Assembly of the State of Rio de Janeiro) and Councilman Reimont — PT (President of the Monitoring Commission on the issue of Smart Cities, in the Municipal Council of Rio de Janeiro).
Parliamentarians who have already joined the initiative:
Benedita da Silva (House of Representatives); Hilton Coelho (Bahia State Legislative Assembly); Laina Pretas por Salvador (Salvador City Council); Renato Roseno (Ceará Legislative Assembly); Gabriel Aguiar (Fortaleza City Council); Fábio Félix (Federal District Legislative Chamber); Camila Valadão (Vitória City Council); Andreia de Jesus (Minas Gerais Legislative Assembly); Bella Gonçalves, Iza Lourença, Macaé Evaristo and Moara Sabóia (City Council of Contagem); Bia Caminha, Enfermeira Nazaré Lima, Lívia Duarte (Belém City Council); Juntas por Pernambuco (Pernambuco Legislative Assembly); Dani Portela, Ivan Moraes (Recife City Council); Carol Dartora (Curitiba City Council); Dani Monteiro (Rio de Janeiro State Legislative Assembly); Benny Briolly, Paulo Eduardo Gomes, Professor Tulio (Niterói City Council); Chico Alencar, Lindbergh Farias, Monica Benício, Paulo Pinheiro, Reimont, Tainá de Paula, Tarcísio Motta, Thaís Ferreira, Wiliam Siri (Rio de Janeiro City Council); Professor Josemar (São Gonçalo City Hall); Bruna Rodrigues, Daiana Santos, Karen Santos, Laura Sito , Matheus Gomes (Porto Alegre City Hall); Coletiva Bem Viver (Florianópolis City Hall); Iran Barbosa (Sergipe State Assembly); Linda Brasil (Aracaju City Hall); Guida Calixto (Campinas City Hall); Flávia Hellen (Paulista City Council); Débora Camilo (Santos City Council); Érica Malunguinho, Isa Penna, Leci Brandão (São Paulo Legislative Assembly); Celso Giannazi, Elaine do Quilombo Periférico, Erika Hilton, Luana Alves, Professor Toninho Vespoli, Silvia da Bancada Ativista (São Paulo City Council), Andreia de Jesus (State Representative of Minas Gerais), Mônica Francisco and Renata Souza (State Representative of Rio de Janeiro).
Translation: Thallita Lima
Revision: Laura Schwartz-Henderson