About me

Hello! I am part of Anthropic’s Frontier Red Team, where I evaluate, improve, and talk about large language models’ (LLMs’) cyber capabilities to help ensure their safe and robust development.

Before joining Anthropic, I completed my PhD at Carnegie Mellon University (CMU) studying under Lujo Bauer in CyLab. My projects involved machine learning for malware detection, cooperative reinforcement learning, and robust machine learning.

Originally from Alaska, I graduated from the US Air Force Academy with a BS in Computer Science and from CMU INI with an MS in Information Security. After serving in several unique roles in the Air Force, I separated as a Captain (O3) to pursue a PhD as an NDSEG Fellow.

Research & Writing

My most recent contributions are on Anthropic’s Frontier Red Team blog, where we regularly post about our work measuring, projecting, and improving frontier LLM capabilities:

• LLM-discovered 0-days - Claude found 500+ high-severity vulnerabilities in open-source projects
• AI Models on Realistic Cyber Ranges - Claude succeeds at multistage network attacks using only standard tools
• AI to Defend Critical Infrastructure - Partnership with PNNL exploring AI-accelerated defense
• AI for Cyber Defenders - Building defensive cybersecurity capabilities
• Claude in Cyber Competitions - Evaluating Claude’s performance in CTF competitions (the subject of my DEF CON talk)
• Cyber Toolkits - An awesome collaboration with Brian Singer on measuring LLMs’ network attack capabilities.

Recent Media Coverage:

• DEF CON 33 Hackers’ Almanack: First featured talk (Feb 2026)
• Axios: Claude Opus 4.6 uncovers 500 zero-day flaws (Feb 2026)
• Axios: PNNL makes key AI cyber defense breakthrough (Jan 2026)
• Fortune: Inside Anthropic’s ‘Red Team’ (Sept 2025)
• Axios: Claude AI takes on human hackers (Aug 2025)