5.7 KiB
IGF 2026 WS #287 AI and security: safeguards in the age of facial recognition
Description
The advance of Artificial Intelligence (AI) has been accompanied by a technosolutionist discourse that links complex problems to overly simplistic responses. In the field of security, this is reflected in the use of Facial Recognition (FR), which has been increasingly deployed in Latin America for purposes ranging from access to commercial buildings and stadiums to the identification of wanted individuals.
In private security, the main concern lies in the indiscriminate use of facial biometrics. Individuals’ consent is not respected, creating data breach risks with serious consequences, since, unlike simple alphanumeric passwords, biometrics cannot be replaced or changed.
In public security, the greatest risk is mass surveillance: people become continuously monitored, without transparency about how their data will be collected, stored, shared, or for how long it will be retained. In this context, socially vulnerable groups are the most affected, as AI systems tend to reproduce existing inequalities. This happens due to underrepresentative datasets and the historical biases embedded in the design and use of these technologies by those who develop them. As a result, these systems make more mistakes with vulnerable groups, generating false positives, abusive police stops, and wrongful arrests that deepen existing inequalities.
This workshop aims to discuss the issue along three main axes: i) the current context of FR use in Latin America; ii) the effects on socially vulnerable groups; and iii) regulatory solutions in the fields of data protection and public security.
It also establishes a direct link to SDG 10, as the workshop shows how the indiscriminate use of AI and FR can deepen existing inequalities and produce disproportionate impacts on socially vulnerable groups, reinforcing the need for measures that promote equity and non-discrimination.
Policy Question(s)
A. What are the main regulatory challenges that arise for public and private security in face of the use of artificial intelligence? B. What is the relation between the use of facial recognition in security and the protection of individuals' personal data? C. How might different social groups be impacted by the effects of AI use on security?
Expected Outcomes
It is intended to leverage the international and multistakeholder dialogue fostered by the workshop to discuss the effects of the use of Artificial Intelligence (AI) and Facial Recognition (FR) in security, in order to formulate possible strategies for policy advocacy, social mobilization, and regulatory action.
Two of the participating organizations, IDEC and TEDIC, maintain active mobilization campaigns on the use of FR in Brazil and Paraguay: “Quem vê cara não vê permissão” and “Con mi cara no”. The workshop aims to use these campaigns as channels to disseminate the discussions held in the workshop, considering their broad and diverse reach and their ability to engage with different audiences and stakeholders. The workshop will seek to produce recommendations for public policymakers on the use of AI in private security. One highlight will be the participation of a speaker from Nairobi, who will comment on the issue from a local perspective.
Format
Roundtable Duration (minutes): 90 Format description: This session format is the best suited because it allows a round of questions to the speakers and a moment for exchange with both the in-person and remote audience, thus ensuring greater engagement and reflection. The workshop will unfold as follows: the in-person moderator will introduce the topic in 12 minutes, commenting on the main themes to be addressed and the objectives. She will pose one question to each of the speakers. Each speaker will have 9 minutes to answer. After that, there will be a 25-minute block for audience interaction, allowing in-person and remote attendees to ask questions and make comments. In the final 5 minutes, the rapporteur will preview the panel’s final output, presenting a set of 5 recommendations for public policymakers on the use of AI in private security, with a focus on the safeguards required for the adoption of facial recognition.
Hybrid Format
Hybrid Format: To ensure a good hybrid experience, we will rely on the support of the online moderator, who will be responsible for facilitating communication between the online speakers and the in-person moderator. Throughout the workshop, in-person and online speakers will have equal speaking time, and during the second block, the audience will be encouraged to interact with all speakers.
Organizers
Organizer 1: Júlia Caldeira, Instituto de defesa de consumidores (idec) Organizer 2: Julia Abad, Instituto de defesa de consumidores (idec) Organizer 3: Lauro Accioly, Universidade Estadual Paulista (UNESP)
Speakers
Speaker 1: J. Carlos Lara, Civil Society, Latin American and Caribbean Group (GRULAC) Speaker 2: Maricarmen Sequera, Civil Society, Latin American and Caribbean Group (GRULAC) Speaker 3: Luiza Dutra, Private Sector, Latin American and Caribbean Group (GRULAC) Speaker 4: Jorge Fontelles, Government, Latin American and Caribbean Group (GRULAC) Speaker 5: Miriam Beartrice Wanjiru, Civil Society, African Group
Moderator
Julia Abad, Civil Society, Latin American and Caribbean Group (GRULAC)
Online Moderator
Júlia Caldeira, Civil Society, Latin American and Caribbean Group (GRULAC)
Rapporteur
Lauro Accioly, Technical Community, Latin American and Caribbean Group (GRULAC)
Participant Benefits
What will participants gain from attending this session?
SDGs
- Reduced Inequalities
- Peace, Justice and Strong Institutions