Jobs/San Francisco/Abuse Investigator (AI Self-Improvement Risk)
San Francisco, California, United States

Abuse Investigator (AI Self-Improvement Risk)

About the Team OpenAI’s mission is to ensure that general-purpose artificial intelligence benefits all of humanity. We believe that achieving our goal requires real world deployment and iteratively updating based on what we learn.

Company
OpenAI
Compensation
$288K - $320K
Schedule
Full-Time
Role overview

What this role actually needs.

Abuse Investigator (AI Self-Improvement Risk) at OpenAI in San Francisco. UpJobz keeps this listing high-signal for applicants targeting serious high-tech roles across the United States, Canada, and Mexico. About the Team OpenAI’s mission is to ensure that general-purpose artificial intelligence benefits all of humanity. We believe that achieving our goal requires real world deployment and iteratively updating based on what we learn.

Responsibilities

Day-to-day expectations

A clear list of the work this role is designed to cover.

  • Review leads, investigate model behavior, and identify cases where systems demonstrate agentic or autonomous patterns that that introduce safety risks
  • Detect and analyze behaviors such as multi-step planning, capability chaining, tool use, persistence, and workaround behavior
  • Develop signals and tracking strategies to help proactively identify emerging agentic risk patterns across our platform
  • Identify gaps in existing safeguards, evaluations, or monitoring systems and propose improvements
  • Communicate investigation findings clearly to technical, policy, and leadership stakeholders
  • Be someone people enjoy working with and appreciate the opportunity to help others
Requirements

What a strong candidate brings

This keeps the job page specific, readable, and easier to match.

    Benefits

    Why people would want this job

    Benefits help searchers understand whether the role is a real fit before they apply.

    • Review leads, investigate model behavior, and identify cases where systems demonstrate agentic or autonomous patterns that that introduce safety risks
    • Detect and analyze behaviors such as multi-step planning, capability chaining, tool use, persistence, and workaround behavior
    • Develop signals and tracking strategies to help proactively identify emerging agentic risk patterns across our platform
    • Identify gaps in existing safeguards, evaluations, or monitoring systems and propose improvements
    • Communicate investigation findings clearly to technical, policy, and leadership stakeholders
    • Be someone people enjoy working with and appreciate the opportunity to help others
    Subscriber playbook

    Turn this listing into an application plan.

    This is the first pass at the premium UpJobz layer: a fast brief that helps serious applicants move with more clarity.

    Next moves

    • Tailor your resume around ai and research instead of sending a generic application.
    • Use the first two bullets of your application to connect your background directly to abuse investigator (ai self-improvement risk) is a high-signal hybrid role in san francisco, and it is most realistic for united states residents.
    • Open the role quickly if it fits and bookmark three similar jobs before you leave the page.

    Interview themes

    Artificial IntelligenceHybridairesearchpythonaws

    Watchouts

    • $288K - $320K is visible, so calibrate your application around the posted range.
    • Use united states residents as part of your positioning so the recruiter does not have to infer it.
    • Show concrete examples of succeeding in hybrid environments.
    SEO context

    Search intent signals for this listing

    Helpful keyword hooks for serious tech searchers and future programmatic job pages.

    Abuse Investigator (AI Self-Improvement Risk)OpenAISan FranciscoUSArtificial Intelligenceairesearchpythonawssecurityplatformapillminfrastructure
    Next step

    Ready to move on this role?

    This page keeps the application flow simple while giving you enough context to decide quickly and move.