How to apply for Security Engineer, Offensive Security

About Anthropic

Anthropic is a frontier AI research and product company working on alignment, policy, and security. It posts specific opportunities it considers high impact, and it notes candidates can read concerns about doing harm by working at a frontier AI company via 80,000 Hours. The security team's stated mission is to safeguard AI systems and maintain user and societal trust.

About the role

This is a Security Engineer, Offensive Security role on Anthropic's Security Engineering team, working remotely in the US. The work covers red and purple team engagements against cloud infrastructure, endpoints, and bare metal deployments, plus penetration testing of high value deployments. The role also involves contributing to AI-assisted security testing tooling and working with other teams on AI-specific attack scenarios.

A typical day

A typical day likely involves running or planning red and purple team engagements against cloud, endpoint, or bare metal targets, and conducting penetration tests on high value deployments. You would also work with other security and engineering teams on AI-specific attack scenarios, contribute to AI-assisted security testing tooling, and document findings for technical and executive audiences. Exact routines are not stated in the job details, so ask about team structure and daily workflow in interviews.

Who Anthropic is looking for

• 5+ years of hands-on red teaming and offensive security operations experience.
• Deep expertise in the areas listed in the requirements (the job details cut off the full list, so confirm the complete requirements on the posting before applying).
• Comfortable simulating advanced threat actors across cloud infrastructure, endpoints, and bare metal.
• Able to document and present findings to both technical and executive audiences, translating attack narratives into risk insights that inform security roadmaps.
• Interested in AI-specific attack scenarios and in building or using AI-assisted security testing tooling and workflows.

Tips for this application

• Read the 80,000 Hours career review on working at an AI lab that Anthropic links in its own company description, and address your reasoning about working at a frontier AI company if relevant.
• Tailor your resume to the exact scope listed: red and purple team engagements, cloud infrastructure, endpoints, bare metal, and penetration testing of high value deployments.
• Give concrete examples of findings you documented and presented to executive audiences, since the role explicitly includes executive communication.
• Show any experience with AI-assisted security testing tooling or AI-specific attack scenarios, because the role calls for both.
• Confirm the full requirements list on the live posting before applying, since the provided job details are truncated mid-requirement.

What to cover in your cover letter

['Your hands-on red team and offensive security work over 5+ years, with specific engagement types and targets.', 'Experience across cloud infrastructure, endpoints, and bare metal environments.', 'How you have translated attack narratives into actionable risk insights for security roadmaps and executive audiences.', "Your interest in AI-specific attack scenarios and AI-assisted security testing workflows, tied to Anthropic's stated security mission."]

Research before applying

• Anthropic's stated mission: reliable, interpretable, and steerable AI systems, and its safety and alignment work.
• The 80,000 Hours career review on working at an AI lab, which Anthropic links directly in its company description.
• The Security Engineering team's stated mission of safeguarding AI systems and maintaining user and societal trust.
• Anthropic's published security or policy material, to understand how the security team fits into the wider company.

Likely interview topics

Based on the job description, expect questions about:

• Walk through a red or purple team engagement you ran against cloud infrastructure, endpoints, or bare metal, and what the outcome was.
• How you would approach a penetration test of a specific high value deployment.
• How you would design or use AI-assisted security testing tooling and workflows.
• How you would simulate advanced threat actors in an environment that includes AI systems.
• How you present attack findings to executive audiences and turn them into risk insights for a security roadmap.

Common mistakes to avoid

• Treating this as a generic security engineering application without addressing red teaming, purple teaming, and penetration testing specifically.
• Ignoring Anthropic's stated mission and the linked reading about working at a frontier AI company.
• Applying without checking the full requirements list, since the provided details are truncated and may omit key skills.