Agentic Safety Policy Lead for Frontier AI

Showcify; United States; On-site; USD 180,000 - 240,000; Full time

Job summary

OpenAI is seeking a researcher to address real-world risks from model misalignment as AI systems scale in autonomy and horizon length. The role focuses on translating vulnerabilities into concrete behavioral policies, evaluations, and safeguards.

Hybrid role based in San Francisco with relocation support. You will work across safety systems, research, engineering, security, and policy to improve frontier AI safety and deployment decisions.

Qualifications

• Strong background in AI safety, privacy, security, or adjacent fields.
• Demonstrated interest in AI alignment and understanding of misaligned model behavior.

Responsibilities

• Identify vulnerabilities as models interact with tools, data, and external systems; translate into safeguards.
• Develop threat models and empirical frameworks for harmful outcomes from misalignment.
• Build frameworks to understand harmful outcomes from model misalignment.
• Identify underlying behaviors and system conditions driving those outcomes.
• Translate findings into policy frameworks, evaluation criteria, and safeguards.
• Develop human data campaigns and gold sets for measurement and evaluation of risks.
• Collaborate with research, engineering, security, product to balance safety, utility and risk.
• Inform deployment decisions, system cards, safeguards reports, and OpenAI’s agentic safety approach.
• Build monitoring approaches to detect regressions and emerging risks post-deployment.

Skills

AI safety

Privacy

Security

Adversarial mindset

AI alignment

Data analysis

Evaluation data

Cross-functional collaboration

Clear communication

Job description

OpenAI is seeking a researcher to address real-world risks from model misalignment as AI systems scale in autonomy and horizon length. The role focuses on translating vulnerabilities into concrete behavioral policies, evaluations, and safeguards.

Hybrid role based in San Francisco with relocation support. You will work across safety systems, research, engineering, security, and policy to improve frontier AI safety and deployment decisions.