Hugging Face Public Policy Program Overview and Submitted Materials
TL;DR
Hugging Face announced a comprehensive, cross‑functional public‑policy program that supplies U.S., EU, and U.K. policymakers with detailed position papers, testimony, and comment letters on AI governance, reflecting the company’s commitment to responsible openness and to shaping regulation from the perspective of its research and developer community.
Overview of Hugging Face’s Policy Workstream
Hugging Face frames its AI policy effort as a multidisciplinary, cross‑organizational workstream rather than a siloed communications or global‑affairs unit. The program draws on expertise from:
- Ethics and Society researchers (e.g., the “Ethics and Society Regulars” series)
- The internal legal team
- Machine‑learning engineers working on domains such as healthcare, art, and model evaluation
The team’s mandate is to translate community‑driven needs on the Hub into responsible‑openness practices, including ethics‑forward research, model‑card transparency, platform safeguards, and policy advocacy.
Policy Materials Shared with Policymakers
Hugging Face maintains a public repository of policy documents that are updated as new submissions are made. The materials are grouped by jurisdiction and include both congressional/legislative requests for information (RFIs) and executive‑level comment letters.
United States
| Category | Date | Document | Focus |
|---|---|---|---|
| Congressional | June 2025 | RFI on the American Science Acceleration Project (ASAP) | AI research acceleration strategy |
| Congressional | Sep 2023 | CEO Clement Delangue – Senate AI Insight Forum kickoff statement | Strategic AI outlook for Senate forum |
| Congressional | Jun 2023 | CEO Clement Delangue – House Committee on Science, Space, and Technology testimony | Written statement and recorded testimony on AI innovation |
| Congressional | Nov 2023 | Chief Ethics Scientist Margaret Mitchell – Senate Insight Forum statement | Ethical considerations for AI policy |
| Executive | Sep 2025 | RFI on NIST AI Standards “Zero Drafts” TEVV | Standards for testing, evaluation, verification, and validation |
| Executive | May 2025 | RFI on the National AI R&D Strategic Plan | Guidance for federal AI research funding |
| Executive | Apr 2025 | RFI on NIST “Zero Draft Pilot Project” | Early‑stage AI standards development |
| Executive | Mar 2025 | RFI on AI Action Plan development | Input on a coordinated U.S. AI action plan |
| Executive | Oct 2024 | Comments on BIS‑2024‑0047 (Reporting requirements for advanced AI models) | Transparency and reporting obligations |
| Executive | Sep 2024 | Response to NIST RFC AI 800‑1 (Managing misuse risk for dual‑use foundational models) | Risk‑mitigation for dual‑use models |
| Executive | Jun 2024 | Response to NIST RFC AI 600‑1 (Generative AI risk‑management framework) | Generative AI risk profile |
| Executive | Mar 2024 | Response to NTIA RFC on dual‑use foundation models with widely available weights | Model‑weight distribution concerns |
| Executive | Feb 2024 | Response to NIST RFI on Executive Order sections 4.1, 4.5, 11 (AI) | Implementation of EO on AI governance |
| Executive | Dec 2023 | Response to OMB RFC on agency AI use | Agency‑level AI deployment guidance |
| Executive | Nov 2023 | Response to U.S. Copyright Office AI & Copyright inquiry | Copyright implications of AI‑generated content |
| Executive | Jun 2023 | Response to NTIA RFC on AI accountability | Accountability frameworks for AI systems |
| Executive | Sep 2022 | Response to NIST AI Risk Management Framework | Early feedback on NIST RMF |
| Executive | Jun 2022 | Response to NAIRR on National AI Research Resource task force findings | Recommendations for national AI research infrastructure |
European Union
| Date | Document | Focus |
|---|---|---|
| Dec 2024 | Open Source Guide to the EU AI Act | Practical guidance for OSS developers under the AI Act |
| Nov 2024 | Comments on the first draft of the EU AI Office’s GPAI Code of Practice | Recommendations for trustworthy general‑purpose AI |
| Sep 2024 | Response to EU AI Office multi‑stakeholder consultation on future‑proof AI Act (trustworthy GAIs) | Survey answers and free‑text submissions |
| Jan 2024 | Response to Digital Services Act transparency‑report requirements | Transparency obligations for AI services |
| Jul 2023 | Comments on the proposed AI Act | Early stakeholder input on EU AI legislation |
United Kingdom
| Date | Document | Focus |
|---|---|---|
| Feb 2025 | Response to UK Intellectual Property Office consultation on copyright and AI | Copyright policy for AI‑generated works |
| Nov 2023 | Oral evidence from Head of Global Policy Irene Solaiman to the House of Lords | Testimony on AI governance and openness |
| Sep 2023 | Response to UK Parliament RFI on large language models (LLMs) | Technical and policy considerations for LLMs |
| Jun 2023 | Response to No 10 AI Regulatory Innovation White Paper RFI | Innovation‑friendly regulatory recommendations |
All documents are hosted in the huggingface/policy-docs dataset, which can be accessed via the Hub for verification and reuse.
Implications for the AI Ecosystem
- Policy influence from the open‑source community – By submitting technically detailed comment letters and testimony, Hugging Face ensures that the perspectives of open‑source model developers are represented in high‑level regulatory drafts.
- Transparency and reproducibility – Publishing the full text of every submission in a public dataset creates an auditable trail, enabling researchers and civil‑society groups to analyze the company’s policy stance.
- Cross‑jurisdictional coordination – The parallel effort across the U.S., EU, and U.K. demonstrates a strategic approach to harmonizing standards, which may reduce regulatory fragmentation for developers who operate globally.
- Embedding ethics in product design – The policy work is tightly linked to internal initiatives such as model‑cards, content‑guidelines, and ethics‑forward research, suggesting that regulatory recommendations will be grounded in concrete technical safeguards.
How to Access the Materials
All policy documents are stored in the huggingface/policy-docs dataset on the Hub. The dataset provides a viewer with version history, file previews, and download links. Researchers can clone the repository or query individual PDFs via the Hub API.
Conclusion
Hugging Face’s public‑policy program systematically channels its multidisciplinary expertise into concrete submissions for U.S., EU, and U.K. policymakers, reinforcing the company’s role as a leading voice for responsible openness in AI governance.
Sources
- OriginalPublic Policy at Hugging Face