Research Engineer, Chemical, Biological and Nuclear AI Safety, DeepMind
Mountain View, CA, USA · Full-time · Posted Oct 13, 2026
- Experience
- 5 years writing code; 3 years training, fine-tuning, or evaluating machine learning models.
- Education
- Bachelor's degree in a Management, Scientific, Technical, or Engineering field, or equivalent practical experience.
About the job
Build and deploy safeguards that evaluate and reduce catastrophic chemical, biological, radiological, nuclear, and explosives risks in frontier AI models while preserving legitimate scientific research.
What you'll do
- Develop safeguards and test them through automated and manual red teaming, measuring impact on legitimate research.
- Design and run risk evaluations for frontier models, including agentic systems, and lead reliability of safeguards and incident response.
- Work with scientific experts to create datasets and automated graders; manage sensitive data and communicate findings, including uncertainty and negative results.
What they ask for
- Bachelor's degree in a Management, Scientific, Technical, or Engineering field, or equivalent practical experience.
- 5 years of experience writing code in a shared, version-controlled codebase with other contributors.
- 3 years of experience training, fine-tuning, or evaluating machine learning models.
Nice to have
- Master's degree or PhD in Chemistry, Physics, Chemical, Nuclear Engineering, Biology, or a Computational Science.
- Experience building classifiers or evaluation infrastructure for large language models; applying statistics to ML evaluation; production software/on-call or incident response; or relevant security, export control, or adversarial work.
Pay
$174000 - $252000 USD base salary
Tools
Gemini
AI focus
Applied AI safety research: evaluate frontier models, including agentic systems, and build, red-team, and launch safeguards against catastrophic scientific misuse.
Model-written summary · the original posting is authoritative