← back
arXivLeihan Zhang, Wecheng Ye, Xianlong Ma, Haochuan Liu, Yang Li, Qianyu Zhang, Jinliang Chen, Qiang YanSat, Jun 6, 2026, 5:06 PM PDT
score 15.8

Large dataset of AI failures and harms from news stories

Original: RiskNet: A large-scale dataset of AI risk incidents from news with alignment and multi-dimensional annotations

Source: arxiv.org

Writing ELI5 summary…