Research Engineer – AI Alignment

About the role

Anthropic is seeking a Research Engineer / Scientist to advance AI alignment research. You will design and execute rigorous machine‑learning experiments aimed at understanding and steering the behavior of powerful AI systems, ensuring they remain helpful, honest, and harmless.

Key responsibilities

• Build and run elegant, thorough ML experiments focused on AI safety and alignment.
• Collaborate with teams in Interpretability, Fine‑Tuning, and Frontier Red Team to test safety techniques.
• Develop and evaluate scalable oversight, AI control, and alignment stress‑testing methods.
• Conduct multi‑agent reinforcement‑learning experiments to probe model robustness.
• Investigate model welfare, moral status, and related safety assessments.

Required profile

• Passion for creating AI that is helpful, honest, and harmless.
• Interest in challenges posed by human‑level AI capabilities.
• Comfortable working as both a scientist and an engineer.
• Willingness to interview in Python and preferably reside in the Bay Area.

Required skills

• Python programming.
• Machine‑learning experimentation.
• Reinforcement learning, including multi‑agent setups.
• Experience with large language models and AI safety research.

What we offer

• Opportunity to work on cutting‑edge AI alignment problems.
• Collaboration with a fast‑growing, interdisciplinary research team.
• Access to state‑of‑the‑art AI resources and infrastructure.