Research Engineer, Responsible Frontier AI Research, DeepMind

Overview

In this role you will prototype scalable safety solutions for the Responsibility research portfolio and design robust training/inference pipelines to detect harmful manipulation in frontier language models. You will develop post-training strategies to mitigate manipulation risks and work with Research Scientists to translate safety research into practical implementations. You will build and maintain evaluation infrastructure to track model safety across releases and present results to cross-functional stakeholders. This role combines hands-on engineering with research-driven safety work at scale, contributing to high-impact AI safety initiatives and broad deployment.

Responsibilities

• Prototype scalable safety solutions across the Responsibility research portfolio
• Architect and optimize training and inference pipelines to detect and evaluate harmful manipulation in language models
• Develop post-training strategies to mitigate deception, sycophancy, and covert influence
• Collaborate with Research Scientists to implement safety research and present results to stakeholders
• Build and maintain evaluation infrastructure to track safety performance across releases

Key requirements

• Master’s degree or PhD in Computer Science or related technical field
• Experience in AI safety or responsible AI research
• Preferred: ML-focused advanced degree (Master’s or PhD) in CS, Engineering, or related field
• Experience in Python and C++ for high-performance ML library development
• Experience with LLM demitigation, adversarial robustness, harmful manipulation detection, or AI safety evaluation/mitigation
• Experience building evaluation frameworks, benchmarks, or automated testing pipelines for ML models
• Prior experience in AI safety or responsible AI research

• collaboration
• cross-functional communication
• problem solving and initiative
• Python
• C++
• high-performance ML library development

…