Research Engineer, Cybersecurity RL

Responsibilities

• Partner with researchers and safety teams to understand analytical needs and build solutions.
• Develop agentic integrations that allow AI systems to autonomously investigate and act on analytical findings.
• Design and implement reinforcement learning environments.
• Conduct experiments and evaluations and deliver work into production training runs.
• Contribute to strategic decisions about what to build, partnerships, and investment priorities.
• Collaborate with researchers, engineers, and cybersecurity specialists across and outside Anthropic.
• Participate in an on-call rotation when required.

Requirements

• Experience with machine learning.
• Experience in cybersecurity research.
• Strong software engineering skills.
• Ability to balance research exploration with engineering implementation.
• Bachelor’s degree or an equivalent combination of education, training, and/or experience.
• Relevant field of study demonstrated through coursework, training, or professional experience.
• Preferred qualifications include professional experience in security engineering, fuzzing, detection and response, or other applied defensive work.
• Preferred qualifications include experience participating in or building CTF competitions and cyber ranges.
• Preferred qualifications include academic cybersecurity research or other experimental and research experience.
• Preferred qualifications include familiarity with reinforcement learning techniques and environments.
• Preferred qualifications include familiarity with LLM training methodologies.

Benefits

• Anthropic sponsors visas and will make every reasonable effort to obtain a visa if an offer is made, though sponsorship is not successful for every role or candidate.
• Competitive compensation and benefits.
• Optional equity donation matching.
• Generous vacation and parental leave.
• Flexible working hours.
• Lovely office space for collaboration.
• Hybrid work policy requiring staff to be in an Anthropic office at least 25% of the time, with some roles requiring more office time.