
Product Manager, Frontier Risks Safety
OpenAI · San Francisco, California, United States
Applications are made on OpenAI's own site. We never sit in the middle.
Job description
About the Team The Frontier Risks Safety team sets out safety requirements and executes planning for frontier model launches, including biological and cyber risks and government models. We work across research, engineering, and policy to assess risks, define and validate safeguards, and drive safety readiness throughout the model development and deployment process. We also help shape frontier trusted-access programs, ensuring that access to advanced capabilities is supported by appropriate safeguards, clear requirements, and rigorous evaluation. About the Role As a Product Manager on the Frontier Risks Safety team, you will own product strategy, requirements, and delivery for initiatives that mitigate frontier AI risks and enable responsible access to advanced capabilities. You will partner with researchers, engineers, policy teams, and independent experts to identify the most important problems to solve, translate research and risk assessments into product requirements, and prioritize what to build. You will guide work from discovery through deployment and iteration, making tradeoffs explicit and measuring whether the resulting safeguards and workflows address the risks they are intended to mitigate. We’re looking for people who combine strong technical judgment, experience building and shipping complex technical products, and a commitment to AI safety. You should be comfortable working through uncertainty, engaging deeply with research, and adapting direction as capabilities, evidence, and user needs evolve. This role is based in San Francisco, CA. We use a hybrid work model of 3 days in the office per week and offer relocation assistance to new employees. In this role, you will: - Set product strategy and priorities. Develop a shared understanding of frontier risks, deployment needs, and gaps in safeguards. Own the roadmap for mitigations, prioritizing investments based on risk, evidence, technical feasibility, and the value the underlying capabilities enable for users. - Drive frontier trusted-access programs. Ensure defenders and scientists have access to the frontier capabilities they need to strengthen security and accelerate scientific discovery. - Turn safety needs into product requirements. Work with model teams, safety researchers, engineers, and policy partners to translate threat models, evaluation findings, and user needs into clear requirements, acceptance criteria, and measurable outcomes. - Drive safeguards from concept to deployment. Partner with research and engineering teams to turn promising safety approaches into effective mitigations across models and products. Resolve product and technical tradeoffs, manage dependencies, and drive delivery of improvements that close material safety gaps. - Own risk-area safety readiness. Own risk assessments and safeguard plans for launches in your area. Define the requirements and evidence needed for readiness reviews with technical partners, surface unresolved risks early, and bring clear recommendations to accountable safety and deployment decision-makers. - Use evaluations to guide product decisions. Partner with researchers to prioritize capability evals, alignment and control assessments, adversarial testing, and safeguard stress-testing, including for autonomous and multi-agent systems. Assess what the evidence establishes, where uncertainty remains, and how findings should change the roadmap. - Align teams and communicate decisions. Build alignment across research, engineering, product, and policy on goals, priorities, and tradeoffs. Resolve blockers, establish clear ownership, and communicate results, limitations, and remaining risks through decision documents, launch reviews, and contributions to system cards and assessment reports. You might thrive in this role if you: - Have 5+ years of product management experience, or equivalent experience owning the direction and delivery of complex technical products. - Have defined and shipped products or systems with researchers and engineers, particularly in AI, cybersecurity, biosafety, or another field where failures can have serious consequences. - Can turn ambiguous risks and user needs into a clear product strategy, prioritized roadmap, actionable requirements, and measurable outcomes. - Understand frontier AI safety concepts, including capability evals, alignment, control, monitoring, and safeguard robustness, and can engage substantively with technical teams. - Can assess what eval results do and do not establish, including methodological limitations, uncertainty, and the difference between model capability and real-world risk. - Exercise sound judgment when evidence is concerning or incomplete, make practical tradeoffs, and escalate unresolved risks constructively. - Have worked extensively across research, engineering, and product teams and can build alignment with policy and communications partners. - Communicate complex technical wor
Other jobs you might want
OpenAI
Client Partner, Ads Solutions – Scaled
New York, United States
OpenAI
Client Partner, New Business - Ads Solutions
New York, United States
OpenAI
Data Science Leader, B2B Marketing & Demand Generation
San Francisco, California, United States
OpenAI
Designer, OpenAI.com
United States
OpenAI
Risk & Fraud Analyst
San Francisco, California, United States
OpenAI
Certification Assessment Specialist
San Francisco, California, United States