Remote AI Safety Research Fellow - Part-Time

Scale AI, Inc.; San Francisco (CA); Remote; USD 114,000 - 162,000; Part time

Job summary

Scale AI, Inc. is seeking researchers for the Human Frontier Collective program to tackle empirical AI safety questions. You will work under a mentor and match to a proposal, meeting weekly with a Scale researcher to guide day-to-day work.

The role is a remote, part-time internship in the United States, offering 20 hours per week for 3 months, with compensation of $100/hour. You will contribute to safety research across control, alignment, multi-agent safety, RSI safety, and evaluation science.

Qualifications

• Pursuing or completed PhD in CS/ML/Statistics or related field.
• Ability to design and run rigorous safety research.
• Strong communication of complex results.

Responsibilities

• Own research end to end: turn safety questions into hypotheses and experiments.
• Build measurement: evaluation harnesses, agent environments, datasets, logs.
• Exercise research judgment: identify confounds and interpret results.
• Make the work legible: produce reproducible code, logs, and papers.

Skills

PhD in CS/ML

Research skills

Strong communication

Confidential handling

Education

PhD in CS/ML/Statistics

Job description

Scale AI, Inc. is seeking researchers for the Human Frontier Collective program to tackle empirical AI safety questions. You will work under a mentor and match to a proposal, meeting weekly with a Scale researcher to guide day-to-day work.

The role is a remote, part-time internship in the United States, offering 20 hours per week for 3 months, with compensation of $100/hour. You will contribute to safety research across control, alignment, multi-agent safety, RSI safety, and evaluation science.