Remote AI Safety Research Scholar

Scale; United States; Remote; USD 34,000 - 55,000; Part time

Job summary

Scale in partnership with the Human Frontier Collective invites researchers to join a small cohort focused on empirical AI safety. This remote internship, based in the United States, offers mentorship and hands-on work on safety evaluation and testing of AI systems.

Scholars will own substantial parts of safety projects and interact with a Scale mentor throughout the process. Selected researchers turn safety questions into testable hypotheses, design controlled experiments, and produce

Qualifications

• Pursuing or has completed a PhD in CS, ML, statistics, mathematics, or a related field.
• Can advance with high-level guidance, including implementing, debugging, and interpreting results.
• Able to handle confidential information and internal systems responsibly.

Responsibilities

• Turn high-level safety questions into hypotheses, baselines, and controlled experiments.
• Work with a Scale researcher to define direction and receive feedback.
• Run experiments end-to-end and produce reproducible results.
• Build measurement harnesses, datasets, and evaluation pipelines.
• Prepare precise technical writing and outputs suitable for top ML conferences.

Skills

Research communication

Independent work

Education

PhD in CS/ML/Stats/Math

Job description

Scale in partnership with the Human Frontier Collective invites researchers to join a small cohort focused on empirical AI safety. This remote internship, based in the United States, offers mentorship and hands-on work on safety evaluation and testing of AI systems.

Scholars will own substantial parts of safety projects and interact with a Scale mentor throughout the process. Selected researchers turn safety questions into testable hypotheses, design controlled experiments, and produce