At the AIAL, we have a comprehensive view of AI accountability. Some of the most stagnant issues we currently encounter are rooted in extractive technological ecologies and oppressive capitalist structures.
Our inquiries into AI accountability, therefore, span from studies of large systems, structures, and ecologies (such as the AI field itself and regulatory processes) to executions of audits and evaluations on specific AI models, tools, and training datasets. We are also invested in conceptual and critical work that advances frameworks and theories of change that underpin algorithmic audit, model evaluation, and meaningful accountability.
We recognize that AI accountability research is most impactful when it can inform the public, impacted groups, and policy makers. Thus, we aim for active policy translation of our (as well as field wide) research.
Recent Publications
-
Computer-vision research powers surveillance technology
Pratyusha Ria Kalluri, William Agnew, Myra Cheng, Kentrell Owens, Luca Soldaini, Abeba Birhane
Nature, 2025. DOI: 10.1038/s41586-025-08972-6 (open-access)
Associated Media:
- News and Views: Computer-vision research is hiding its role in creating ‘Big Brother’ technologies
- Video: Is AI powering Big Brother? Surveillance research is on the rise
- News: Wake up call for AI: computer-vision research increasingly used for surveillance
- Editorial: Don’t sleepwalk from computer-vision research into surveillance
Community Work
The AIAL has supported and signed the following community initiatives and efforts:
-
No AI For Atrocity Crimes
with AI scholars, scientists, human rights advocates and experts on 9 Jul 2025 -
Open Joint Letter against the Delaying and Reopening of the AI Act
with Center for Democracy & Technology (CDT), European Consumer Organisation (BEUC), European Digital Rights (EDRi), European Centre for Not-for-Profit Law (ECNL) on 9 Jul 2025