Whitzard AI

Whitzard AI is a research team at Fudan University dedicated to AI safety and security. Our research focuses on conversational LLMs, multimodal foundation models, and AI agents, with an emphasis on safety evaluation, red-teaming, trustworthy generation, and governance-oriented methods.

In this organization, we release open datasets, benchmarks, models, and tools developed around JADE, our large-model safety evaluation and governance platform. We aim to support transparent, reproducible, and practical AI safety research for both the academic community and real-world applications.