Security Researcher: AI Threats & Agentic Risks

Rozgoo; United States; Remote; USD 293,000 - 405,000; Full time

Job summary

OpenAI is seeking a deeply technical security researcher to help prepare for security threats from advanced AI agents. You will identify potential compromise paths and design long-lead security controls, then stress-test defenses via AI agent evaluations and penetration tests.

The role requires strong software engineering skills, familiarity with modern infrastructure, and the ability to engage with stakeholders across teams.

Qualifications

• Deep technical security and modern infrastructure knowledge.
• Strong software engineering skills with prototype-building interest.
• Ability to engage with stakeholders effectively.

Responsibilities

• Identify paths by which capable future internal AI agents could compromise OpenAI.
• Design security controls focusing on long lead-time measures.
• Stress-test defenses with AI agent evaluations and penetration tests

Skills

Security engineering

Software engineering

Cloud infrastructure

Containers / CI-CD

Distributed systems

Stakeholder communication

Job description

OpenAI is seeking a deeply technical security researcher to help prepare for security threats from advanced AI agents. You will identify potential compromise paths and design long-lead security controls, then stress-test defenses via AI agent evaluations and penetration tests.

The role requires strong software engineering skills, familiarity with modern infrastructure, and the ability to engage with stakeholders across teams.