Frontier AI Security Researcher & Red Team Lead

Aisi; Greater London; Hybrid; GBP 65.000 - 120.000; Vollzeit

Benefits dieser Stelle

Hybrid working

Central London office

Work-from-home stipend

Generous paid parental leave

Pension contributions

Zusammenfassung

AI Security Institute is hiring for a Research Engineer role in London. You will design and run ML experiments on frontier models, build robust evaluation tooling, and publish findings that influence safety standards.

You’ll work across research and engineering with strong autonomy and high-impact ownership. Join a mission-driven team with frontier access, substantial compute and government-facing collaborations.

Qualifikationen

• Design, build and run ML experiments on frontier models autonomously.
• Write clean, reusable code for ML experiments and model fine-tuning/inference.
• Critically assess how experiments support safety claims and conclusions.
• Demonstrate impact-driven mindset and collaborative work style.

Aufgaben

• Research: design and run ML experiments, gather empirical evidence, design adversarial attacks, publish results.
• Testing: evaluate frontier labs' monitors, model threats, and report implications.
• Build tooling and experimental pipelines for fast, high-quality results.
• Leverage LLMs to automate attack, evaluation and analysis loops.
• Develop and run the infrastructure for training and serving models at scale.

Kenntnisse

ML experimentation

Software engineering

Safety evaluation

Team collaboration

Autonomy

Tools

LLM finetuning frameworks

Inspect

RL optimization

Jobbeschreibung

AI Security Institute is hiring for a Research Engineer role in London. You will design and run ML experiments on frontier models, build robust evaluation tooling, and publish findings that influence safety standards.

You’ll work across research and engineering with strong autonomy and high-impact ownership. Join a mission-driven team with frontier access, substantial compute and government-facing collaborations.