Technical Program Manager, AI Safety & Safeguards

About the role

OpenAI's mission is to ensure that artificial general intelligence benefits all of humanity, and this role exists to translate that mission into executable safeguards across our most critical products and infrastructure. You will own the end-to-end execution of complex, high-stakes initiatives that turn abstract safety commitments into deployed systems with measurable impact on risk reduction. This means shaping technical roadmaps, defining program architecture, and coordinating cross-functional work that spans model development, infrastructure, product, operations, legal, policy, and external partners. You will balance the urgency of deployment with the rigor required to prevent violent misuse and other serious harms while maintaining model usefulness. A core part of your job will be to establish clear ownership of emerging risks and ensure that safety and operational controls are integrated into cloud and API platforms. You will drive the development and rollout of evaluations, classifiers, monitoring systems, abuse detection, enforcement workflows, and agent-assisted review where appropriate. You will also define and own metrics that track program delivery, deployment readiness, safety effectiveness, and user or developer experience. Through all of this, you will communicate technical tradeoffs and program decisions to senior leadership while advocating for both users and developers.

Key facts

What you'll do

Partner with product, engineering, research, and design to shape technical roadmaps, long-term platform direction, and safeguards priorities across ChatGPT, API, enterprise, and cloud environments.

Translate product and safety objectives into execution plans that balance near-term delivery, long-term platform evolution, customer needs, and responsible deployment.

Lead cross-functional programs end to end, from problem definition and technical design through implementation, launch, iteration, and operational follow-through.

Own milestones, dependencies, risk mitigation, decision-making, and accountability across engineering, infrastructure, integrity, trust and safety, legal, policy, operations, and go-to-market teams.

Partner with engineers on system architecture, public and internal APIs, cloud deployment patterns, platform integrations, operational failure modes, and technical tradeoffs.

Drive the development, integration, or rollout of safeguards such as model evaluations, post-training mitigations, classifiers, abuse detection, monitoring, enforcement workflows, human review, and agent-assisted review.

Coordinate readiness for sensitive or high-impact deployments, ensuring appropriate safety coverage, escalation paths, operational controls, and alignment with external cloud or enterprise partners.

Identify emerging misuse and severe-harm risks; define ownership and response mechanisms across prevention, detection, investigation, enforcement, and continuous improvement.

Establish meaningful metrics for program delivery, deployment readiness, safety effectiveness, operational quality, and user or developer experience.

Represent customer and developer needs in technical and leadership decisions, balancing risk reduction, model usefulness, implementation complexity, speed, and scale.

Communicate program health, risks, recommendations, and consequential tradeoffs clearly to engineering teams, cross-functional partners, and executives.

Improve delivery and safety outcomes through practical operating mechanisms, scalable tooling, and data- or AI-assisted workflows.

Requirements

You must have a strong track record of delivering high-impact technical programs across product, platform, API, infrastructure, machine learning, or safety domains.

You must bring strong technical judgment across distributed systems, cloud deployments, public or internal APIs, developer platforms, production machine learning, or scalable product architecture.

You must be able to engage credibly with engineers on system design, implementation tradeoffs, reliability, deployment constraints, and operational risks.

You must have delivered platforms or developer-facing capabilities and understand how technical decisions affect customer experience, developer experience, and product usefulness.

You must understand how safety, integrity, or abuse-prevention systems work, including evaluations, model guardrails, classifiers, abuse detection, monitoring, enforcement, human review, and agent-assisted review.

You must be comfortable navigating ambiguity, balancing safety with model usefulness, and advocating for users and developers in complex environments.

You must be able to lead cross-functional work with urgency, rigor, and empathy while communicating clearly to both technical and executive audiences.

You must have a record of owning end-to-end accountability for programs, including milestones, dependencies, risk mitigation, and decision-making across multiple teams.

Nice to have

Only roles and experiences that are explicitly mentioned in the source are included under this section, and the source does not enumerate any preferred qualifications, backgrounds, or skills beyond the requirements listed above.

Practical notes

This is a full_time position based in San Francisco. The role involves significant cross-functional coordination and may require travel to support deployment readiness reviews, partner meetings, and operational reviews with cloud and enterprise teams. No specific visa sponsorship details, hours beyond standard full-time expectations, or application deadlines are provided in the source.