Frontier AI Safety Research Engineer

Google Inc.; Greater London; On-site; GBP 70,000 - 100,000; Full time

Job summary

Google Inc. is hiring for a role focused on AI safety mitigations in London. The successful candidate will develop classifiers to detect misuse, evaluate AI systems against threats, and apply machine learning techniques to ensure safety and ethics in technology. Requirements include a Bachelor's degree, substantial experience in software development, and knowledge of cybersecurity principles. Join a collaborative environment at Google DeepMind to help build safe AI systems while pushing the boundaries of technology for public benefit.

Qualifications

• 5 years of experience with software development in various programming languages.
• Experience testing, maintaining, or launching software products.
• 1 year of experience in software design and architecture.

Responsibilities

• Develop advanced classifiers and data pipelines to detect misuse.
• Create cross-context monitoring systems to identify attack vectors.
• Implement account-level response systems to track malicious actors.
• Evaluate and secure agentic AI systems through threat modeling.
• Advance research in adversarial robustness and automated red-teaming.

Skills

Software development

Testing and maintaining software products

Software design and architecture

Cybersecurity detection and response

Machine learning

Automated red-teaming

Model interpretability

Architectural judgment

Education

Bachelor’s degree or equivalent

PhD in Computer Science or related field

Job description

Google Inc. is hiring for a role focused on AI safety mitigations in London. The successful candidate will develop classifiers to detect misuse, evaluate AI systems against threats, and apply machine learning techniques to ensure safety and ethics in technology. Requirements include a Bachelor's degree, substantial experience in software development, and knowledge of cybersecurity principles. Join a collaborative environment at Google DeepMind to help build safe AI systems while pushing the boundaries of technology for public benefit.