Product Manager, Safety Systems

Details

About the team

Safety Systems manages the complete lifecycle of safety efforts for OpenAI’s frontier models, ensuring models are deployed responsibly and have a positive impact on society. The team's work spans research and engineering initiatives — from system-level safeguards and model training to evaluation and red-teaming — all aimed at mitigating misuse and maintaining a high bar for safety. The team leads OpenAI's commitment to developing and deploying safe Artificial General Intelligence (AGI), fostering trust, responsibility, and transparency. The group focuses on learning from deployments, distributing AI's benefits widely, and ensuring powerful tools remain aligned with human values and safety considerations.

About the role

As a Product Manager on the Safety Systems team you will drive initiatives to ensure OpenAI’s frontier model deployments are safe, impactful, and aligned with user needs and technical innovation. You will clarify strategic priorities, develop safety-focused product roadmaps, and collaborate closely with AI researchers, software engineers, policy experts, and cross-functional partners. This role suits a proactive, technically skilled product manager adept at adversarial thinking and excited to tackle ambiguous problems through structured analysis and collaborative decision-making. The position is based in San Francisco, CA, with relocation assistance available.

Responsibilities

• Partner closely with research, engineering, and data science teams, as well as policy stakeholders, to embed safety throughout development and deployment of frontier AI models.
• Develop comprehensive frameworks for understanding and mitigating deployment safety risks, drawing on data analysis, expert consultation, and adversarial assessments.
• Define strategic priorities and product roadmaps focused on improving deployment safety, enhancing reliability, and managing emerging AI capabilities.
• Create scalable methodologies, tools, and processes for evaluating, refining, and continuously improving safety systems.
• Establish repeatable processes to integrate cutting-edge AI safety research into models and product offerings.
• Develop and continuously refine clear, actionable metrics that capture safety performance and user experience at scale.

Requirements

• 6+ years of product management or related industry experience, with specific expertise in AI safety, trust & safety, integrity, or related domains.
• Strong cross-functional collaboration skills and the ability to communicate complex technical concepts clearly and persuasively to research, engineering, policy, and product partners.
• Demonstrated ability to drive consensus and action in ambiguous spaces and to identify and challenge underlying assumptions through insightful questioning.
• Comfort with adversarial thinking, data-informed risk assessment, and developing frameworks and metrics to measure safety and reliability.
• Interest in interdisciplinary fields such as human-computer interaction, psychology, philosophy, or similar areas is a plus.

Benefits

• Competitive base pay within the listed range, plus equity and potential performance-related bonuses.
• Medical, dental, and vision insurance with employer contributions to Health Savings Accounts.
• Pre-tax accounts for Health FSA, Dependent Care FSA, and commuter expenses.
• 401(k) retirement plan with employer match.
• Generous parental, medical, and caregiver leave policies.
• Paid time off (flexible PTO for exempt employees), paid company holidays, and paid sick/safe time as required by law.
• Mental health and wellness support, employer-paid life and disability coverage, annual learning and development stipend, daily meals in offices (or meal credits), and additional fringe benefits.
• Relocation support for eligible employees.

Other notes

• OpenAI is an equal opportunity employer and provides reasonable accommodations for applicants with disabilities. Background checks will be administered in accordance with applicable law. Contact and policy links are provided in the original posting.