Research Engineer, Rule of Law

TL;DR

Research Engineer, Rule of Law (AI safety): Conduct technical and sociotechnical research at the intersection of AI and constitutional democratic institutions, with an accent on legal-alignment safety evaluations, fine-tuning, and institutional vulnerability analysis. Focus on developing novel AI applications for civic life and translating research into actionable recommendations for product and policy.

Company

Anthropic builds reliable, interpretable, and steerable AI systems and advances research that supports safe and beneficial AI for users and society.

What you will do

• Design and run AI safety evaluations focused on legal alignment, and perform technical work (including fine-tuning) to improve targeted model performance.
• Use AI to analyze institutional vulnerabilities created or exacerbated by AI and develop sociotechnical mitigations.
• Develop novel AI applications to bolster and enrich democratic processes.
• Run rigorous sociotechnical studies on how AI-provided information supports informed deliberation and debate.
• Collaborate closely with researchers, policy and law experts, and cross-functional teams across Anthropic.
• Translate research insights into actionable recommendations for both product and policy.

Requirements

• Deep expertise in AI plus substantive expertise in government, law, political science, or public policy.
• Advanced deep learning skills and capabilities in data science, mechanism design, govtech, deliberative tech, or a related field.
• Strong experimental design, data analysis, and inferential statistics skills.
• At least five years of work experience in academia, industry, or government.
• Ability to navigate ambiguity in novel research settings where objectives are hard to define.
• Strong sociotechnical skills: understanding human factors and how AI systems interact with groups of humans.

Nice to have

• Experience designing and running evaluations or fine-tuning experiments on large language models.
• Large-scale data analysis experience.
• UX and interface design experience for intuitive tools for non-technical users.
• Engineering with privacy considerations in mind.
• Research related to deliberative technology, civic technology, or societal impacts of AI.

Culture & Benefits

• Location-based hybrid policy: expect staff to be in an office at least 25% of the time.
• Visa sponsorship available; not guaranteed for every role/candidate.
• Generous vacation and parental leave, flexible working hours, and optional equity donation matching.
• Collaborative research environment with frequent research discussions and emphasis on communication skills.

Hiring process

• Application review with consideration of qualifications even if not all requirements are met.
• Interviews focused on research/technical fit and collaboration across safety, policy, and product.
• Use only verified Anthropic recruiter email domains during the process.

Location: San Francisco, CA or New York City, NY (hybrid; at least 25% in-office time)

Salary: $320,000 - $485,000 USD annually