Research Engineer, Frontier Safety Mitigations, DeepMind
Mountain View, CA, USA · Full-time · Posted Oct 12, 2026
- Experience
- 5 years in software development; 3 years testing, maintaining, or launching software products; and 1 year in software design and architecture.
- Education
- Bachelor’s degree or equivalent practical experience.
About the job
Build and deploy end-to-end safety mitigations for frontier AI models, including systems that detect misuse, coordinated harms, and attacks on agentic AI.
What you'll do
- Build misuse classifiers and data pipelines, from automated evaluation through model iteration.
- Develop cross-session monitoring and account-level response systems to detect and address persistent malicious actors.
- Threat-model, test, and mitigate agentic AI risks, and advance automated red-teaming and adversarial robustness.
What they ask for
- Bachelor’s degree or equivalent practical experience.
- 5 years of experience with software development in one or more programming languages.
- 3 years of experience testing, maintaining, or launching software products.
- 1 year of experience with software design and architecture.
Nice to have
- Master's degree or PhD in Computer Science or a related technical field.
- Experience in LLM development, fine-tuning, or safety evaluation; taking research from concept to product; or Safety and Alignment.
Pay
$174000 - $252000 USD base salary
Tools
Large Language Model (LLM)
AI focus
Research and implement safety mitigations for frontier models, including misuse detection, agentic AI security, automated red-teaming, and adversarial robustness.
Model-written summary · the original posting is authoritative