AI Safety Researcher

About the Role

DeepMind is looking for a Research Scientist to join our Technical AI Safety team. Your goal will be to develop methods for ensuring that increasingly capable AI systems remain under human control and behave according to human values. You will conduct original research on topics such as scalable oversight, mechanistic interpretability, and robust alignment, publishing your findings at top-tier conferences.

Requirements

• PhD in CS, Math, or Philosophy
• Strong background in machine learning theory
• Experience with interpretability tools
• Excellent mathematical reasoning
• Passion for long-term AI safety

Benefits

• World-class research environment
• Competitive compensation
• Generous pension scheme
• Private healthcare
• On-site restaurant