The AI Verification and Evaluation Research Institute (AVERI) is a US-based nonprofit with the mission to make third-party auditing of frontier AI effective and universal. AVERI aims to bring about a world in which the most powerful AI systems – and the companies that build them – are rigorously audited for safety and security by third parties. We believe this independent auditing layer is essential for enabling confident deployment of AI and managing critical risks from this increasingly powerful technology.
About the role:As a Research Scientist at AVERI, you will conduct research on frontier AI systems and the safety and security practices surrounding them, often as part of pilot audits with frontier companies, sometimes working externally to those companies and sometimes embedded inside them. The work sits at the practical and scientific foundations of assessing model capabilities, risks, robustness, and alignment-relevant behavior in ways that are technically deep, decision-relevant, and useful to the broader AI governance ecosystem. You will work closely with our policy, engineering, and standards teams, and you will help shape a new field as it develops. We are hiring at two levels: Research Scientist and Research Director (see below). Given AVERI’s early stage (launched in January 2026), early staff take on a wide range of responsibilities.
What You’ll Do:Own a line of research on frontier AI evaluation and auditing, from scoping through delivery, within AVERI's research agenda.
Design and run evaluations of closed and open-weight AI systems, including as part of pilot audits with frontier AI companies.
Develop audit methods and evaluation protocols for model behavior, capabilities, safety properties, robustness, deployment risks, or compliance with stated policies.
Build reproducible experimental pipelines for testing models, analyzing outputs, and interpreting results.
Produce rigorous audit reports and translate technical findings into actionable recommendations for AI developers, policymakers, auditors, and the broader public.
Publish and present completed research as conference papers, white papers, open-source tools, talks, and public writing.
Collaborate with researchers, engineers, policy experts, and standards bodies inside and outside AVERI, and contribute to research planning and methodological standards for frontier AI audits.
Handle confidential access to company systems and practices and sensitive findings with appropriate care.
A track record of independent research in AI/ML, preferably in AI safety, evaluations, robustness, or security (roughly three or more years beyond a PhD or equivalent, or a comparable record of delivered work).
Strong technical judgment: you can design rigorous empirical studies and reason carefully about measurement, uncertainty, and evidence.
Experience working with large language models or other modern AI systems, especially in safety-relevant contexts.
You can turn open-ended questions into tractable research plans and communicate clearly with collaborators while working independently in a remote environment.
You are motivated by AI safety, AI governance, and third-party auditing, and want to contribute to it full-time.
Prior work on AI auditing, model evaluations, red teaming, interpretability, robustness, privacy, cybersecurity, or responsible AI.
Published research or high-quality public technical writing, and experience building reproducible evaluation pipelines, benchmarks, or research infrastructure.
Familiarity with AI governance, standards, regulatory processes, or institutional audit practices, and experience communicating technical findings to non-technical audiences or working in interdisciplinary teams.
Domain knowledge in areas relevant to AI risk, such as cybersecurity, biosecurity, misinformation, privacy, legal compliance, or model misuse.
We are also open to hiring this role at the Research Director level. The Research Director owns and sets AVERI’s technical research agenda, priorities, and roadmap. They would also represent AVERI’s research externally to frontier labs, standards bodies, policymakers, and the research community, and would help to recruit a larger research team (which they may manage directly or just provide technical leadership to). We expect a Research Director to bring roughly seven or more years in the field, an exceptional profile, a publication record peers recognize, and a demonstrated record of setting research direction that others followed. If you think you might be a strong fit at this level, please say so in your application.
Someone with several years of independent research experience who will lead their own line of work will typically come in as a Research Scientist. Level and title are set during the interview process.
What we offer:Salary: Your salary depends on the scope, autonomy, and impact we expect you to have at AVERI. The salary ranges for this role are:
Research Scientist: $220,000 to $450,000
Research Director: $330,000 to $600,000
Comprehensive medical benefits, including generous reimbursement of eligible HRA expenses
6% unrestricted retirement contribution
Unlimited PTO
A work enablement and professional development package of up to $22,200 per year
A mission-driven team committed to integrity and purpose, providing a supportive environment with meaningful opportunities for career growth
Location and travel: Fully remote, with a preference for candidates in the San Francisco Bay Area. Travel at least twice per quarter and possibly significantly more, especially to Washington, DC and San Francisco.
If you think you are perfect for this role, but you do not meet all of the requirements, we strongly encourage you to apply and tell us why you’d be a great fit. We are committed to fostering a diverse and inclusive environment where all employees have the opportunity to succeed. We do not discriminate based on race, religion, color, national origin, gender, sexual orientation, or any other legally protected status. Please let us know what accommodations you might need when applying or if asked to be interviewed by emailing us at [email protected]
