Critical Harm Operations Analyst

Rozgoo; San Francisco (CA); Hybrid; USD 150,000 - 210,000; Full time

Job summary

OpenAI in San Francisco seeks experienced Trust & Safety operators to investigate complex cases, apply usage policy, and build scalable systems to reduce risk. You will act as SME on high-stakes escalations, coordinating with Legal, Compliance, Security, Product, and Engineering for defensible outcomes.

This role involves designing processes, tooling, and automation for safe operations at scale, handling sensitive content, and turning ambiguity into clear, repeatable workflows.

Qualifications

• Experience in Trust & Safety and policy enforcement.
• Strong analytical and data-driven decision making.
• Ability to collaborate with Product, Engineering, Legal, and Compliance.

Responsibilities

• Apply usage policy with rigor and nuance on complex scenarios.
• Mitigate material harm and escalate high-severity content cases.
• Serve as escalation SME for high-stakes incidents with cross-functional teams.
• Design scalable trust workflows and QA processes.
• Drive automation and tooling to improve speed and accuracy.

Skills

Trust & Safety

Policy enforcement

Data analysis

Cross-functional collaboration

LLM tooling

Job description

About the Team

Trust & Safety Operations is central to protecting OpenAI’s platform, customers, and the public from abuse. We support a diverse customer base -- from individual users and early-stage startups to global enterprises -- across ChatGPT, our API, and new product surfaces as they launch. Within the Support organization, we partner closely with Product, Engineering, Legal, Policy, Go To Market, and Operations teams to deliver a great user experience at scale while reducing material harm and mitigating catastrophic risks.

About the Role

We’re hiring experienced Trust & Safety / Critical Harm operators who can investigate complex cases, apply and evolve usage policy in real-world scenarios, and help build scalable systems that reduce risk over time. You will contribute as a subject-matter expert (SME) on high-stakes escalations, partnering with cross-functional stakeholders to drive fast, defensible outcomes. You will also help design the processes, tooling, and automation that power safe operations at scale. This role is ideal for someone who combines strong judgment with sharp analytical instincts, and who enjoys turning ambiguity into clear decisions, repeatable workflows, and durable automation. Please note: This role may involve handling sensitive content, including material that may be highly confidential, sexual, violent, or otherwise disturbing. Location: San Francisco, CA (hybrid: 3 days in office/week).

In this role you will:

• Apply usage policy with rigor and nuance: Interpret and apply OpenAI’s usage policies to complex, novel scenarios; provide clear guidance to customers and internal teams; document edge cases and propose policy refinements.
• Mitigate material harm and catastrophic risks: Triage, assess, and support actions on content and behavior that can drive real-world harm, including high-severity domains; escal…
• Serve as an escalation SME for high-stakes cases: Support incident response and executive-visible escalations by producing clear assessments, recommending next steps, and coordinating with Legal, Compliance, Security, Product, and Engineering as needed.
• Build scalable trust workflows: Design and operate processes for human-in-the-loop labeling, content/user reporting, appeals, enforcement actions, and continuous QA -- with a high bar for quality and consistency.
• Drive automation and operational efficiency: Identify repeatable patterns, translate them into requirements, and partner with Engineering and Data teams to ship tooling and automation (including LLM-enabled automation) that improves speed, accuracy, and coverage.
• Analyze trends and strengthen feedback loops: Use quantitative and qualitative analysis to surface emerging abuse patterns, measure policy and tooling performance, and feed insights back into detection systems, product mitigations, and policy updates.
• Raise the quality bar: Define and monitor KPIs, build calibration and QA programs, iterate on reviewer training, and improve guidelines and tooling based on error analysis.
• Enable internal and external teams: Create playbooks, SOPs, and training that help partner teams understand our enforcement posture, risk thresholds, and operational philosophy.

You might thrive in this role if you:

• Build for scale: You’ve taken workflows from zero to one and then scaled them without sacrificing quality.
• Bring deep Trust & Safety experience: 5+ years in Trust & Safety, integrity, risk, policy enforcement; experience working with vendors is a plus.
• Have strong judgment under ambiguity: You can make defensible calls in gray areas, write clearly, and adjust quickly as new information arrives.
• Are analytically strong: You can assess risk, spot trends, and use data to prioritize problems and evaluate solutions (data fluency is a plus).
• Bias toward automation: You’ve shipped operational efficiencies through tooling, process redesign, and automation; you’re comfortable leveraging LLMs to improve triage, labeling, QA, or enforcement consistency.
• Operate well cross-functionally: You can translate nuanced operational reality to engineers and policy stakeholders, and drive alignment without escalation-by-default.
• Stay humble and collaborative: You learn quickly, share context generously, and optimize for team outcomes.
• Have experience with high‑severity safety domains (for example: CBRN, cyber abuse).
• Have experience building QA programs, calibration loops, and measurable reviewer performance systems.
• Have hands‑on experience writing requirements for internal tools, piloting automation, or partnering closely with Engineering on safety systems.

About OpenAI

OpenAI is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence benefits all of humanity. We push the boundaries of the capabilities of AI systems and seek to safely deploy them to the world through our products. AI is an extremely powerful tool that must be created