Research Scientist, AI Safety and Alignment

The role

You will conduct research and build evaluations, tools, and datasets to understand and reduce risks from frontier AI systems. This includes designing control protocols, building alignment evaluations, and studying monitorability.

What you will do

• Design and stress-test control protocols and monitoring systems
• Build alignment evaluations for misaligned behaviors
• Study monitorability and how training methods affect it
• Investigate risks from multiple AI agents interacting
• Develop threat models and evaluations for automated AI research
• Work with government agencies and labs to design evaluations
• Publish research methodologies and write technical reports

Required qualifications

• Commitment to the mission of promoting safe, secure, and trustworthy AI
• Practical experience conducting technical research collaboratively
• Comfort building and instrumenting ML pipelines
• Comfort building evaluation harnesses
• Comfort running experiments on frontier models and agents
• Ability to turn research literature into working prototypes
• Familiarity with current research in at least one of the mentioned areas
• Track record of published research in machine learning, particularly in AI safety and alignment
• At least three years of experience working on sophisticated ML problems in research or product development
• Strong written and verbal communication skills

Nice to have

• Experience designing evaluations and benchmarks
• Experience building tools and pipelines to analyze model and agent trajectories at scale
• Experience with red-teaming or adversarial testing of AI systems
• Familiarity with AI safety policy and governance frameworks

Benefits

• Base salary
• Equity
• Comprehensive health, dental and vision coverage
• Retirement benefits
• Learning and development stipend
• Generous PTO
• Commuter stipend

Before you apply

• Research interviews assess practical ML prototyping, debugging, research concepts, and organizational culture alignment
• 90-day waiting period before reconsidering candidates for the same role

About the company

Scale is a data and evaluation partner for frontier AI companies. Their mission is to develop reliable AI systems for the world's most important decisions.