AI Risk Categorization Decoded (AIR 2024)
We present a comprehensive AI risk taxonomy derived from eight government poli- cies from the European Union, United States, and China and 16 company policies worldwide, making a significant step towards establishing a unified language for generative AI safety evaluation. We identify 314 unique risk categories, organized into a four-tiered taxonomy. At the highest level, this taxonomy encompasses System & Operational Risks, Content Safety Risks, Societal Risks, and Legal & Rights Risks. The taxonomy establishes connections between various descriptions and approaches to risk, highlighting the overlaps and discrepancies between public and private sector conceptions of risk. By providing this unified framework, we aim to advance AI safety through information sharing across sectors and the promotion of best practices in risk mitigation for generative AI models and systems.
Authors
- Yang Yu (ORCID: https://orcid.org/0000-0002-6708-9885)
- Percy Liang (ORCID: https://orcid.org/0000-0002-0458-6139)
- Ruoxi Jia (ORCID: https://orcid.org/0000-0001-9662-9556)
- Kevin Klyman (ORCID: https://orcid.org/0009-0003-2130-3529)
- Bo Li (ORCID: https://orcid.org/0000-0003-4186-4828)
- Yi Zeng (ORCID: https://orcid.org/0000-0002-8820-2488)
- Dawn Song
- Minzhou Pan
- Andy Zhou
Institutions
- University of Illinois Urbana-Champaign (US)
- University of California, Los Angeles (US)
- Harvard University Press (US)
- University of Chicago (US)
- Eastern University (BD)
- Virginia Tech (US)
- University of California, Berkeley (US)
- Stanford University (US)
Publication Details
- Journal
- SuperIntelligence - Robotics - Safety & Alignment
- Published
- 2024-10-09
- DOI
- https://doi.org/10.70777/si.v1i1.10603
- Citations
- 9
- Primary Topic
- Ethics and Social Impacts of AI
- Type
- article
- Field-Weighted Citation Impact
- 4.73