AI Risk Categorization Decoded (AIR 2024): From Government Regulations to Corporate Policies
Abstract
We present a comprehensive AI risk taxonomy derived from eight government policies 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.
Keywords
Cite
@article{arxiv.2406.17864,
title = {AI Risk Categorization Decoded (AIR 2024): From Government Regulations to Corporate Policies},
author = {Yi Zeng and Kevin Klyman and Andy Zhou and Yu Yang and Minzhou Pan and Ruoxi Jia and Dawn Song and Percy Liang and Bo Li},
journal= {arXiv preprint arXiv:2406.17864},
year = {2024}
}