Responsible AI in the OSS: Reconciling Innovation with Risk Assessment and Disclosure
Open Access
- 15 October 2025
- journal article
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society
- Vol. 8 (1) , 513-527
Abstract
Ethical concerns around AI have increased emphasis on model auditing and reporting requirements. We thoroughly review the current state of governance and evaluation practices to identify specific challenges to responsible AI development in OSS. We then analyze OSS projects to understand if model evaluation is associated with safety assessments, through documentation of limitations, biases, and other risks. Our analysis of 7902 Hugging Face projects found that while risk documentation is strongly associated with evaluation practices, high performers from the platform’s largest competitive leaderboard (N=789) were less accountable. Recognizing these delicate tensions from performance incentives may guide providers in revisiting the objectives of evaluation and legal scholars in formulating platform interventions and policies that balance innovation and responsibility.Keywords
All Related Versions
This publication has 25 references indexed in Scilit:
- On the suitability of hugging face hub for empirical studiesEmpirical Software Engineering, 2025
- Systematic analysis of 32,111 AI model cards characterizes documentation practice in AINature Machine Intelligence, 2024
- Deep Learning Model Reuse in the HuggingFace Community: Challenges, Benefit and Trends2024
- HFCommunity: An extraction process and relational database to analyze Hugging Face Hub dataScience of Computer Programming, 2024
- Auditing large language models: a three-layered approachAI and Ethics, 2023
- Governing AI safety through independent auditsNature Machine Intelligence, 2021
- Deep learning: a statistical viewpointActa Numerica, 2021
- The global landscape of AI ethics guidelinesNature Machine Intelligence, 2019
- The reusable holdout: Preserving validity in adaptive data analysisScience, 2015
- Gradient-based learning applied to document recognitionProceedings of the IEEE, 1998