Research Engineer / Scientist, Frontier Red Team (Cyber)

About this role

Anthropic's Frontier Red Team seeks a Research Engineer/Scientist to develop AI-powered cybersecurity defenses and evaluate advanced AI cyber capabilities. You'll build tools and frameworks to protect against AI-enabled threats, design experiments with autonomous systems, and produce policy-informing research on frontier AI risks in cyberspace.

What you'll do

• Develop systems and tools for AI-powered vulnerability discovery, malware detection, and network security
• Design and conduct experiments to evaluate autonomous AI capabilities in realistic cybersecurity scenarios
• Build infrastructure for testing AI systems in security-focused environments
• Create technical demonstrations and artifacts for policymakers and public stakeholders
• Collaborate with external cybersecurity and national security experts to validate research
• Lead research strategy, roadmap definition, and cross-functional partnerships (senior candidates)

What they're looking for

• Cybersecurity and security research expertise
• Python software engineering
• LLM-based agents and autonomous systems research
• Experiment design and rapid iteration
• Offensive security, vulnerability research, or exploit development
• Cross-functional leadership and communication
• Technical problem-solving and end-to-end ownership
• Discretion and integrity with sensitive work

Benefits

• Relocation support (SF-based role)
• Work on high-impact AI safety research with real-world policy influence
• Collaboration with leading AI safety and cybersecurity experts
• Opportunity to shape Anthropic's cyberdefense research program (senior roles)
• Access to frontier AI systems and research infrastructure