Current Position & Biography
Nina da Hora is a Brazilian computer scientist, founder of Instituto da Hora and researcher working on algorithmic justice. Her public work connects the technical evaluation of AI systems with digital rights, education and questions about who has influence over computing infrastructure.
Her author biography describes research on facial-recognition and computer-vision systems, including their performance on Black faces. Through Instituto da Hora, she brings research and educational work into contact with communities that have often been underrepresented in technology development. She also communicates computing to wider audiences through writing, video and audio projects.
Notable Contributions
The ACM FAccT 2025 program listed da Hora’s CRAFT session “Decoding Bias: Navigating Ethnic Diversity in Facial Recognition Technologies.” Its proposed discussion examined how facial geometry, deep-learning systems and ethnicity interact, and how those relationships can affect recognition accuracy and bias. The program documents a research and discussion contribution, not a finding that every recognition system behaves identically.
Her public-facing work includes Computação sem Caô and the Ogunhê podcast, which discusses the history of African scientists. These projects extend her interest in participation beyond professional AI research into the way computing and scientific knowledge are explained.
In an August 2026 Canaltech column, she argued that open hardware and the RISC-V instruction-set architecture offer Brazil an opportunity to develop greater technological autonomy. She also emphasized a limitation: open designs do not remove dependence on the fabrication capacity needed to manufacture chips. Her argument links digital sovereignty to sustained technical expertise rather than treating openness as a complete substitute for industrial infrastructure.
Why Katharos Is Following Them
Da Hora brings a Brazilian perspective to the intersection of AI auditing, digital rights and infrastructure policy. Her work connects questions about model performance with the institutions, educational opportunities and design decisions that shape who can build and challenge technology.
Areas to Follow
Facial-recognition evaluation; algorithmic justice; technical education; open hardware; and digital-sovereignty policy in Brazil.
Reporting Scope
Coverage focuses on published research, documented programs and clearly attributed policy arguments. Advocacy positions are distinguished from regulatory decisions, and proposed methods from independently validated results. Technical performance claims require evidence about the particular system and evaluation conditions.
