Cyber Red Team Engineer - AI Safety & Adversarial Testing

OpenAI; San Francisco (CA); Hybrid; USD 210,000 - 270,000; Full time

Benefits offered by this job

Relocation assistance

Hybrid work model

Job summary

OpenAI is seeking a Red Teamer focused on cyber to evaluate model capabilities and safeguards across secure testing environments. You will design rigorous experiments, conduct hands-on testing, and translate results into actionable risk assessments for Security, Research, Product, Policy, and Engineering partners.

The role requires strong cybersecurity or model-evaluation background, with capability to build testing tools and collaborate across teams.

Qualifications

• Cybersecurity expertise with hands-on testing and risk assessment.
• Experience evaluating AI model behavior and safeguards at scale.
• Ability to write code and build practical testing tools.
• Comfort working across technical and non-technical teams to drive decisions.

Responsibilities

• Design and run rigorous evaluations of model cyber capabilities and safeguards.
• Conduct hands-on testing to understand model use by security practitioners.
• Distinguish policy and benchmark failures from real-world risk.
• Build automated testing infrastructure for repeatable measurement and analysis.
• Test novel abuse risks in agentic systems.
• Translate findings into risk assessments and actionable recommendations.
• Contribute time to Safety Bug Bounty work.

Skills

Cybersecurity

Model evaluation

Programming skills

Cross-functional collaboration

Job description

OpenAI is seeking a Red Teamer focused on cyber to evaluate model capabilities and safeguards across secure testing environments. You will design rigorous experiments, conduct hands-on testing, and translate results into actionable risk assessments for Security, Research, Product, Policy, and Engineering partners.

The role requires strong cybersecurity or model-evaluation background, with capability to build testing tools and collaborate across teams.