OpenAI
Software Engineer, AI Safety
$207K–$385K
San Francisco · Hybrid · Full-time · Posted Mar 17, 2025
About the job
Build trust and safety systems that detect and prevent platform abuse, improve AI model alignment, and address incidents and system failures.
What you'll do
- Architect, build, and maintain anti-abuse and content moderation infrastructure.
- Collaborate with engineers and researchers to use AI techniques to measure, monitor, and improve model alignment with human values.
- Diagnose and remediate platform incidents, and develop tools and infrastructure to address root causes.
What they ask for
- Experience building and operating production services in a rapidly scaling environment.
- Ability to debug live issues and restore systems quickly.
- Experience with Python or a modern language such as C++, Rust, or Go, with ability to ramp up on Python.
- Strong project management skills; self-directed and able to drive projects with minimal guidance.
Nice to have
- Experience in content safety, fraud, or abuse, or motivation to work on present-day AI safety.
- Experience deploying classifiers or machine learning models, or interest in learning modern ML infrastructure.
Pay
$207K – $385K USD base salary
Tools
Terraform, Kubernetes, Azure, Python, Postgres, Kafka, C++, Rust
AI focus
Use AI techniques to measure, monitor, and improve model alignment with human values, and support safe deployment and abuse prevention.
Model-written summary · the original posting is authoritative