Director of AI Alignment

Job Summary

Leading the alignment and interpretability research agenda for security-domain AI, the full-time remote Director of AI Alignment will develop methods to explain model behavior, detect misuse signals, and establish evaluation frameworks to ensure models operate within intended bounds.

Key responsibilities:

• Own and prioritize the alignment and interpretability research agenda, addressing complex open problems in security-domain AI
• Build techniques for detecting offensive misuse signals in model internals and collaborate with the adversarial evaluation team
• Develop alignment methodologies and evaluation frameworks that ensure models operate safely and effectively

Required qualifications:

• MS or PhD in machine learning, computer science, or a related field with research depth in interpretability or AI alignment
• 8+ years of experience in ML research or engineering, specifically in interpretability or alignment research on large language models
• Hands-on expertise with mechanistic interpretability methods applied to real models
• Experience designing and conducting alignment evaluations that support meaningful safety claims
• Proven track record of leading and developing research teams while actively contributing technically