Beyond English: Assessing Generative AI Risks with Korean Culture
A new dataset and risk framework, AssurAI, tackles the crucial need for culturally nuanced safety evaluations of generative AI models, moving beyond the limitations of English-centric benchmarks.








