Victoria Westerhoff is a Principal AI Security Researcher at Microsoft, where she leads operations for Microsoft's AI Red Team—the group responsible for adversarially testing high-risk generative AI systems, Copilots, agents, and frontier models before they ship. Tori’s team of white-hat hackers works across security, responsible AI, societal, and frontier AI harm domains, helping identify vulnerabilities ranging from traditional cyber risks to emerging challenges such as autonomous behavior, social engineering, and misuse of advanced AI systems.
With a background spanning cognitive neuroscience, national security consulting, strategy, and AI, Tori brings an interdisciplinary perspective to AI security. In addition to standing up the frontier and emerging risks AI Red Team functions, she has contributed to industry initiatives including the MITRE ATLAS framework, Microsoft's open-source PyRIT red teaming toolkit, and Frontier Governance Framework while collaborating with government, academic, and industry partners on frontier AI evaluation and governance.
A recognized thought leader in AI safety and security, Tori has spoken at conferences and institutions including Black Hat USA, RSAC, Microsoft Build, Princeton, MIT, ISU, Georgetown CSET, and international AI Safety Institute forums. She was named one of the Top 100 Women in AI and is a frequent commentator on podcasts and international reporting on the future of AI security, adversarial testing, and responsible AI deployment.