Latin American Conference on AI Ethics 2026 (Niteroi)

LAAI-Ethics 2026 is the second Latin American Conference on Ethics in Artificial Intelligence, held from 16 to 18 September 2026 at the Instituto de Computação of the Universidade Federal Fluminense (UFF), Campus Praia Vermelha, in Niterói across the bay from Rio de Janeiro. It is an academic conference with peer-reviewed proceedings, not a policy summit or a trade show, and its distinguishing feature is that it deliberately builds the ethics agenda from a Latin American and Caribbean starting point rather than importing one.

The organisers’ framing is explicit about why that matters. Their case is that the shared historical trajectories of Latin American countries create problems — and require solutions — that frameworks written elsewhere do not address, while still recognising that ethical concerns in AI are universal and that work from other regions can contribute. The conference therefore pulls in participants from outside the region as well, and mixes disciplines on purpose: computer scientists sit alongside communications researchers, health and legal professionals, and philosophers. A stated goal is building local capacity to understand and implement ethical AI practice, not just to critique it.

Topics of interest span evaluation methods and practices for AI systems; the development and deployment of safe and ethical systems; historical, humanistic, social and cultural perspectives on AI ethics; critical studies of AI-based technologies; laws, regulation, policy and governance of AI systems; and philosophical approaches to questions of agency and moral and epistemic responsibility. Across all of them, contributions grounded in the cultures, norms, laws and sociopolitical systems of Latin America and the Caribbean are explicitly prioritised.

Two keynotes are confirmed. Dr Márjory Da Costa-Abreu, Associate Professor in Ethical AI at Sheffield Hallam University and theme lead for “Responsible, Ethical and Life-Centred AI” at its Centre of Excellence for AI and Robotics, speaks on “Childhood for sale? Profit-Driven AI, the ECA Digital, and our duty to protect the next generation” — a talk on the tension between profit-driven recommender systems, chatbots, games and biometric technologies and children’s rights, read through Brazil’s new ECA Digital legislation and its provisions on age verification, security by default, targeted advertising and content moderation. Anápuàka Muniz Tupinambá Hãhãhãe — Indigenous communicator, journalist and founder of Rádio Yandê, Brazil’s first Indigenous web radio station, and originator of the concept of Indigenous Ethnomedia — speaks on “Indigenous AI: Community Protocols, Data Sovereignty, and Decolonial Ethics for the Global South”, arguing that Indigenous peoples should be treated as knowledge producers and system architects rather than research subjects, and examining the limits of universal AI ethics frameworks when they overlook coloniality, algorithmic racism, data extraction and epistemicide.

The reviewing process is worth noting because it practises what the conference preaches. Papers are up to five pages in SBC format — four pages of main text plus a mandatory one-page Ethical Considerations section — submitted anonymously through Microsoft CMT, and may be written in Portuguese, Spanish or English. Every submission must carry an Ethical Considerations statement describing the ethical challenges the authors faced and how they mitigated them, plus a Generative AI Usage statement; LAAI prohibits the use of large language models to generate text for publications, and failure to disclose the use of generative AI tools is grounds for desk rejection. Accepted papers are indexed with a DOI and published in the proceedings on the SBC OpenLib portal, with at least one author required to register and present.

The 2026 cycle is well advanced: submissions closed on 22 May, final decisions were issued on 29 June and camera-ready versions were due on 6 July, with a registration deadline of 10 July. Depending on volume, the organisers note that not every accepted paper is guaranteed an oral slot — some will be presented as posters. General chairs are Mariza Ferro (IC-UFF), Dárlinton Carvalho (UFSJ) and Livy Real (Kunumi / IComp-UFAM), with programme chairs Sandra Avila (IC-UNICAMP) and Marisol Marini (USP); the programme committee draws on UNICAMP, USP, UFMG, UFRJ, Fiocruz, IFMG, Tecnológico de Monterrey and Sheffield, and there is a dedicated Equity and Inclusion committee. The organisers have negotiated discounted rates at a Niterói hotel for participants.

Worth attending if you work on AI governance, algorithmic fairness, data sovereignty or the sociology of AI and want the regional debate at first hand, in a venue small enough that the conversations continue outside the room.