Beyond AI Collective
gives guidance to avoid discriminatory AI systems.
The Beyond AI Collective is a non-profit association that works to prevent discrimination mediated by the use of algorithmic systems.
"Artificial intelligence" (AI) is a technology whose use can lead to or reinforce discrimination. Public institutions and key private actors are increasingly using AI systems in key areas of life. Often, they lack resources and know-how to put ethical and legal anti-discrimination requirements into practice. We want to close this gap.
value-based, socio-technical and open
You’re using an AI system or in the process of developing one? We support you in assessing and reducing discrimination risks in algorithmic systems. To do this, we analyze AI systems with a holistic approach in which we consider the used data, AI model(s) and training methods as well as the application context. The application context includes, for example, the embedding of outputs in decision-making structures as well as complaint and evaluation processes. On this basis, we develop appropriate measures to mitigate discrimination risks and support implementation. Our work ranges from statistical analyses and monitoring to support for inclusion, diversity and organizational processes. Our approach is defined by a continuous eye-level collaboration with you and your stakeholders.
Along the typical AI development and deployment cycle, we map risks and mitigation measures in the "AI Discrimination Mitigation Framework" in a scientifically sound and practice-oriented way. Together with our network, we continuously expand the framework and make it openly available.
civic, interdisciplinary and intersectional
Our actions are carried out with a civil society perspective. We want to strengthen structurally disadvantaged groups and anti-discrimination work in the digital age and make both more visible. We dedicate ourselves to this in an interdisciplinary, intersectional way and in close association with our growing network. In addition to the concrete analysis and monitoring of AI systems, our agenda therefore includes education and research work. We also represent the civil society perspective in position papers, advocacy and networking.
Team / Contact persons
Founded in 2022 on a volunteer basis, the Beyond AI Collective brings together diverse professional backgrounds and perspectives. Our contact persons:
Stina Lohmüller is a policy advisor and cultural scientist specializing in digital cultures and philosophy of technology. She works at the intersection of digital policy, social justice, and ethics and focuses, among other things, on the operationalization of ethical values and the value-based development of AI systems.
Tobias Sterbak is a Data Scientist and software developer in the field of machine learning and natural language processing (NLP). He works as a freelancer and develops and improves machine learning systems in a variety of application areas.
Michael Puntschuh is a freelance researcher on human rights in the digital age. He has analyzed sources of algorithmically mediated discrimination, contributed to the development of an AI trust standard, and coordinated the Algo.Rules project, which developed process criteria for the design of algorithmic systems.
Tim Vallée is a lawyer and political scientist. He is an expert on the regulation of the digital world and focuses on AI, discrimination risks, the platform economy, sustainability and the common good. He is also a lecturer at the law department of the University of Bonn and has founded KLuG - Köln leben & gestalten e.V., an association that promotes cooperative urban development based on civil society.
Carolina Reis is a lawyer and policy advisor. She works at the intersection of democracy, human rights, and technology, advocating for a bright, democratic, and inclusive present in the digital age.