Institution profile

Weizenbaum Institute for the Networked Society

Academic institutioneurope · de
Official website
Research library6linked papers
Opportunities0open roles
Selected work

Representative Papers

AI Narrative Breakdown. A Critical Assessment of Power and Promise

Jun 23, 2025Conference on Fairness, Accountability and Transparency

This study addresses how dominant AI narratives often obscure the underlying political dimensions, power structures, and value-laden assumptions embedded in artificial intelligence, leading to misjudgments of its societal impacts and gaps in governance. Focusing on the discourse surrounding ChatGPT’s release, the project draws on an interdisciplinary synthesis of critical computer science, Science and Technology Studies (STS), data protection theory, semiotics, and philosophy of mind to systematically deconstruct the prevailing utopian/dystopian binary. It introduces the concept of “Zeitgeist AI” to uncover the implicit power relations and normative choices shaping AI development, emphasizing AI’s nature as a human-directed tool requiring robust social regulation. The work ultimately advocates for a more grounded, responsible AI discourse and governance framework attuned to these socio-technical complexities.

4 citationsRead paper

On the (im)possibility of sustainable artificial intelligence. Why it does not make sense to move faster when heading the wrong way

Mar 22, 2025

This paper critically examines the feasibility of artificial intelligence (AI) as a tool for sustainable development, arguing that AI exacerbates Global North–South inequities across three dimensions—material resource consumption, data extraction, and sociopolitical power asymmetries—while its purported technical neutrality obscures underlying political choices. Method: Drawing on Science and Technology Studies (STS), critical data studies, and transformative sustainability science, the study develops the novel analytical framework of “AI’s triple materiality” and conducts interdisciplinary qualitative critical analysis. Contribution/Results: It advances the principles of “de-datafication” and “small is beautiful,” integrating digital degrowth, anti-extractivism, and public interest theory into AI critique. The paper demonstrates that “sustainable AI” is inherently unattainable under current paradigms and warns against techno-solutionism’s obfuscation of systemic transformation. It advocates democratically grounded, pre-emptive governance to redefine AI’s role and boundaries within sustainability transitions.

1 citationsRead paper

The Benchmark Trap: Structures of Power and Injustice in AI Evaluations

Aug 15, 2026

This study addresses the exacerbation of power concentration and structural inequities by AI benchmarks through an innovative integration of oppression theory into evaluative critique. Synthesizing social theory, critical technology studies, and institutional discourse analysis, it reconstructs the sociopolitical dimensions of benchmarking. The research elucidates how benchmarks, as sociotechnical artifacts, entrench power asymmetries and constrain research trajectories, thereby revealing the systemic harms perpetuated by "benchmark culture." Transcending purely technical evaluation paradigms, this work warns that such practices undermine both cognitive robustness and socially beneficial progress in AI. Ultimately, it provides essential theoretical foundations for establishing equitable AI governance frameworks.

0 citationsRead paper
Recent publications

Latest Papers

The Benchmark Trap: Structures of Power and Injustice in AI Evaluations

Aug 15, 2026

This study addresses the exacerbation of power concentration and structural inequities by AI benchmarks through an innovative integration of oppression theory into evaluative critique. Synthesizing social theory, critical technology studies, and institutional discourse analysis, it reconstructs the sociopolitical dimensions of benchmarking. The research elucidates how benchmarks, as sociotechnical artifacts, entrench power asymmetries and constrain research trajectories, thereby revealing the systemic harms perpetuated by "benchmark culture." Transcending purely technical evaluation paradigms, this work warns that such practices undermine both cognitive robustness and socially beneficial progress in AI. Ultimately, it provides essential theoretical foundations for establishing equitable AI governance frameworks.

0 citationsRead paper

Rule of Thumb: Explaining Artificial Intelligence Systems using Partial Information

Aug 11, 2026

This work addresses the challenge of explaining AI decisions in complex scenarios—such as model-agnostic settings, zero-shot learning, and scientific discovery—by proposing a “Rule of Thumb” (RoT) approach. RoT generates concise, intuitive local explanations by identifying the most critical features for a given prediction. It introduces partial information modeling into explainable AI (XAI) for the first time, enabling effective auditing of black-box models, interpretation of zero-shot large language models, and support for scientific discovery, all while aligning with prevailing AI regulations. The method employs a model-agnostic feature importance assessment coupled with efficient computational strategies, achieving high fidelity without sacrificing speed. To facilitate reproducibility and real-world deployment, the authors have open-sourced the implementation.

0 citationsRead paper

AI Narrative Breakdown. A Critical Assessment of Power and Promise

Jun 23, 2025Conference on Fairness, Accountability and Transparency

This study addresses how dominant AI narratives often obscure the underlying political dimensions, power structures, and value-laden assumptions embedded in artificial intelligence, leading to misjudgments of its societal impacts and gaps in governance. Focusing on the discourse surrounding ChatGPT’s release, the project draws on an interdisciplinary synthesis of critical computer science, Science and Technology Studies (STS), data protection theory, semiotics, and philosophy of mind to systematically deconstruct the prevailing utopian/dystopian binary. It introduces the concept of “Zeitgeist AI” to uncover the implicit power relations and normative choices shaping AI development, emphasizing AI’s nature as a human-directed tool requiring robust social regulation. The work ultimately advocates for a more grounded, responsible AI discourse and governance framework attuned to these socio-technical complexities.

4 citationsRead paper