Lecturer in Artificial Intelligence,
Government & Policy at the Oxford Internet Institute (University of Oxford).
This is me
I'm a Departmental Research Lecturer in AI, Government & Policy at the Oxford Internet Institute (OII). I investigate how datafication is transforming social, economic, and political worlds. Building on my experience as a mathematician and computer scientist, my interest lies in investigating risks and limits of algorithmic governance from a transdisciplinary perspective. My interests lies in analysing the political and environmental impact of AI beyond the code: from mineral extractivism, to labour exploitation and electronic waste dumps.
As a postdoctoral researcher at King’s College London (UK), I explored how datafication technologies are transforming borders and migration governance. I developed digital methodologies to discover which algorithmic systems, such as biometric systems, are used in the field of border security. My transdisciplinary work stems from the ability to collaborate with scholars from a range of different disciplines, including political science, philosophy and law. In 2022, I received the Post-Doctoral Enrichment Award by The Alan Turing Institute.
In the past, I conducted research in transdiciplinary teams to design technological solutions in the private sector. I am a former fellow of the Data Science for Social Good program at University of Chicago. Nowadays, I collaborate with the Jevon’s Paradox blog, where we examine the relationship between science and technology, knowledge and power. I am a research advisor on Algorace, a project that raises awareness about the algorithmic risks, harms and limitations on racialised subjects. I am also part of the Algorights, a civil society initiative, including volunteers and community-based organizations, that analyses AI's impacts on society from an ethical perspective.
Machine Learning Digital methods
Natural Language Processing
Fairness and Accountability AI Ethics
Social Justice Power and Resistance
Valdivia, A., and Tazzioli, M. (2023). Datafication Genealogies beyond Algorithmic Fairness: Making Up of Racialised Subjects. In 2023 ACM Conference on Fairness, Accountability, and Transparency.
Valdivia, A., et al. (2022). There is an elephant in the room: Towards a critique on the use of fairness in biometrics. AI and Ethics, 1-16.
Valdivia, A. et al. (2022). Neither opaque nor transparent: A transdisciplinary methodology to investigate datafication at the EU borders. Big Data & Society 9.2: 20539517221124586.
Valdivia, A. et al. (2022). Judging the algorithm: A case study on the risk assessment tool for gender-based violence implemented in the Basque country. arXiv preprint arXiv:2203.03723.
Valdivia, A. et al. (2021). How fair can we go in machine learning? Assessing the boundaries of accuracy and fairness. International Journal of Intelligent Systems, 36(4), 1619-1643.
Omnes et singulatim: Collective and individual subjectivities in algorithmic governmentality
This workshop aims to inaugurate a (trans)disciplinary debate about how to critically rethink subjectivities in AI.
Speakers: Antoinette Rouvroy, Claudia Aradau, Bernard Harcourt, Seda Gürses, Colin Koopman, Lorena Jaume-Palasí.
Organised by: Daniele Lorenzini (Warwick University), Martina Tazzioli (Goldsmiths), Ana Valdivia (King's College).
Examining the impact of the lifecycle of algorithms
Artificial intelligence (AI) has a wide political and environmental impact. In the last years academics, journalists and civil organisations have centred their efforts in analysing algorithmic discriminatory codes and bias; yet little is known about the impact of the whole lifecycle of AI: from mineral extractivism, to data infrastructures and electronic waste dumps.
Colonial Legacies, Biometric Futures: from Galton to the Entry-Exit System
I was invited to the seminar organised in the frame of the MA in International Relations (Goldsmiths, University of London) to present my work on the theoretical foundations of biometrics and how it has evolved through the development of computational methods such as deep learning.