AI Safety Policy Lead for Frontier Models

Society for AI Security; San Francisco (CA); Hybrid; USD 170,000 - 250,000; Full time

Job summary

OpenAI is seeking a safety-focused role on the Safety Systems Model Policy team in San Francisco to address real-world risks from model misalignment. You will translate insights into behavioral policies, evaluations, monitoring, and safeguards, working with research, engineering, security, product, and policy functions.

This hybrid role is based in San Francisco with relocation support and a three-day-in-office schedule plus optional remote work on Thursdays and Fridays.

Qualifications

• Brings strong background in AI safety, privacy or adjacent fields.
• Understanding of misaligned model behavior and safety policies.
• Ability to work with evaluation data and identify patterns.

Responsibilities

• Identify vulnerabilities as models interact with tools and data, translate into safeguards.
• Develop threat models and empirical frameworks for harmful outcomes.
• Turn findings into policy frameworks, evaluations, and safeguards.
• Collaborate with research, engineering, security, product, and policy teams.
• Inform deployment decisions and safety documentation.

Skills

AI safety

Adversarial thinking

Evaluation of models

Data analysis

Tools

Python

Job description

OpenAI is seeking a safety-focused role on the Safety Systems Model Policy team in San Francisco to address real-world risks from model misalignment. You will translate insights into behavioral policies, evaluations, monitoring, and safeguards, working with research, engineering, security, product, and policy functions.

This hybrid role is based in San Francisco with relocation support and a three-day-in-office schedule plus optional remote work on Thursdays and Fridays.