CAS:
Co-Auditing AI with Civil Society, For Civil Society
Building an AI Audit Resource with and for Civil Society Organisations (CSOs)

Our Project.
Responsible Researchers: David Hartmann, Goda KlumbytÄ—
CAS aims to pursue the participatory development of a transdisciplinary AI auditing & investigation resource for civil society organisations (CSOs). In doing so, this project addresses some of the key challenges CSOs face when analysing AI: opacity, lack of access to data and models, and limited technical knowledge & organisational resources.
By co-developing a digital auditing resource, we aim to support CSOs across the EU in critically examining and contesting the impacts of contemporary AI systems.
The main objective of this project is to provide direct, concrete support to CSOs by co-creating a digital resource for critical auditing of AI systems that CSOs can use in their work. This resource consists of a repository of audit strategies interconnected through tagging, case studies, and tools, and it facilitates information exchange and networking. This repository is available for access to trusted members and collaborators via the login at the top of this page.
Partnered Organisations.

Beyond AI Collective
This section will highlight specific details about a particular service you offer. We will write about what this offering is and how your company uses it to help clients or users achieve their desired goals.

Weizenbaum Institute
This section will highlight specific details about a particular service you offer. We will write about what this offering is and how your company uses it to help clients or users achieve their desired goals.

ApTI
This section will highlight specific details about a particular service you offer. We will write about what this offering is and how your company uses it to help clients or users achieve their desired goals.

Homo Digitalis
This section will highlight specific details about a particular service you offer. We will write about what this offering is and how your company uses it to help clients or users achieve their desired goals.

AI Forensics
This section will highlight specific details about a particular service you offer. We will write about what this offering is and how your company uses it to help clients or users achieve their desired goals.
Key Dimensions of Our Approach: Co-Research
-
Collaboration & Participatory Co-Design
-
Democratic Knowledge Production & Distribution
-
Co-Ownership & Minimising Extractiveness
Our approach centres collaboration in the design process, utilising situated auditing toolkit development (Kraft et al. 2021 & Katell et al. 2020) and a co-research approach (Miceli et al. 2025). While the former highlights the need to situate (Haraway 1988) auditing tools and embed them in community needs, the latter requires addressing CSOs as active co-researchers shaping content and outputs. Following these imperatives, the final digital resource can be situated in CSOs’ practice to address real-world constraints, enabling meaningful use beyond the project timeline.
Co-research is a participatory approach to research where research is conducted by and with people (in this case – organisations) that this research concerns. In this way, the
co-research approach seeks to minimize extractiveness of research and instead espouses democratic and collaborative knowledge production.

Co-research is a participatory approach to research where research is conducted by and with people (in this case – organisations) that this research concerns. In this way, the
co-research approach seeks to minimize extractiveness of research and instead espouses democratic and collaborative knowledge production.
Get In Touch
We welcome inquiries from fellow researchers, journalists, students, NGOs, or other civil society organisations interested in addressing similar challenges relating to AI.