Secure autonomy
Protect the value of agent autonomy by giving it a trustworthy operating boundary.
WHITZARD · 白泽
Security infrastructure for the agentic AI era
Whitzard combines runtime control through AgentGuard, frontier-risk evidence through NUWA, and reusable safety technology through its open ecosystem.
- 01RISK EVIDENCE
- 02CONTROL POLICY
- 03RUNTIME FEEDBACK
- 04OPEN TOOLS
Team vision
Evidence before claims
Ground every claim in public sources, evaluations, and deployment evidence.
Research into control
Translate risk understanding into precise interventions in real systems.
Public safety goods
Build verifiable, reusable models, tools, data, and evaluation infrastructure.
Founding and research team
Meet the people connecting agent security engineering, frontier safety research, and real-world cybersecurity expertise.
Dr. Jiarun Dai
CEOWhitzard CEO and Associate Research Professor at Fudan University, researching binary analysis, vulnerability discovery, and exploitation. Former captain of the Fudan Baize CTF team. He received his PhD from Fudan University under the supervision of Prof. Min Yang.
Dr. Xudong Pan
CTOWhitzard CTO, Associate Research Professor at Fudan University, and PhD advisor at the Shanghai Innovation Institute, researching AI safety, model safety, systems security, and frontier-risk evaluation. He received his PhD from Fudan University under the supervision of Prof. Min Yang.
Dr. Geng Hong
Strategic & Research PartnerAssistant Research Professor at Fudan University researching cybercrime, privacy, and in-the-wild security, collaborating with Whitzard and NUWA. He received his PhD from Fudan University under the supervision of Prof. Min Yang.