AI Safety Researcher — Equity & Sabbatical Perks

Best AI Tools Wiki; San Francisco (CA); Hybrid; USD 180,000 - 240,000; Full time

Benefits offered by this job

Equity stake

Education budget

Relocation support

Sabbatical program

Job summary

Anthropic is seeking an AI Safety Researcher to advance our understanding of AI alignment and develop practical techniques for building safe and beneficial AI systems. You will conduct original research on constitutional AI, interpretability, and scalable oversight.

The role emphasizes independent research, collaboration with a cross-functional team, and a strong publication record to contribute to cutting-edge safety advances at Anthropic.

Qualifications

• PhD in ML/AI Safety or related field.
• Track record of publications in AI safety or alignment.
• Strong mathematical foundations in statistics and optimization.

Responsibilities

• Advance understanding of AI alignment, constitutional AI, and scalable oversight.
• Conduct original research and publish results.
• Collaborate with interdisciplinary teams to deploy safe and beneficial AI systems.

Skills

AI safety research

Python

PyTorch

independent research

Education

PhD in ML/AI Safety

Tools

JAX

Job description

Anthropic is looking for an AI Safety Researcher to advance our understanding of AI alignment and develop practical techniques for building safe and beneficial AI systems. You will conduct original research on constitutional AI, interpretability, and scalable oversight.

Requirements

• PhD in Machine Learning, AI Safety, or related field
• Track record of publications in AI safety or alignment
• Strong mathematical foundations in statistics and optimization
• Proficiency in Python and modern ML frameworks
• Demonstrated ability to conduct independent research

Nice to Have

• Experience with interpretability or mechanistic interpretability
• Background in philosophy of mind or ethics
• Experience with red-teaming or adversarial testing of LLMs
• Familiarity with Constitutional AI or RLHF methods

Significant equity in a high-growth AI company

$10,000 annual conference and education budget

Relocation assistance

Sabbatical program

Skills

AI Safety Alignment Interpretability Python PyTorch Research

This listing was posted on 2026-03-09 and is almost certainly closed. It is kept for reference — check the employer’s own careers page for current openings.

Vincony has all 400+ AI models in one place — compare responses, AI debate, Image/Video/Voice generator, and 20 more tools to help you learn and build with AI.