Role Overview

The Safety Systems team at OpenAI is committed to tackling emerging safety challenges in AI. Drawing from years of practical alignment work, we develop fundamental solutions for safe deployment of advanced models and future AGI. As a Software Engineer in AI Safety, you'll design anti-abuse infrastructure, content moderation tools, and monitoring systems. You'll collaborate across engineering and research to harness AI's potential responsibly.

This role demands strong production engineering skills in high-scale environments. You'll work on our tech stack including Terraform, Kubernetes, Azure, Python, Postgres, and Kafka. Ideal candidates thrive in fast-paced settings, debugging live issues and scaling services amid hyper-growth. If you're passionate about 'now-term' AI safety, fraud prevention, and model alignment, this is your opportunity to shape the future of trustworthy AI.

OpenAI's mission is to ensure general-purpose AI benefits humanity. We prioritize safety at the core, valuing diverse perspectives to build inclusive systems. Located in San Francisco, this on-site role offers unparalleled impact in one of the world's leading AI organizations.

Key Responsibilities

In this pivotal role, you'll:

• Architect, build, and maintain anti-abuse and content moderation infrastructure to safeguard users from unwanted behavior.
• Partner with engineers and researchers to deploy AI techniques measuring and enhancing model alignment to human values.
• Diagnose and remediate live platform incidents, creating tools that address root causes.
• Design scalable systems for fraud detection, content safety, and risk reduction across the platform.
• Deploy ML classifiers and integrate novel safety models into production workflows.
• Assess risks for new features/products, devising innovative mitigations that preserve user experience.
• Scale production services using Kubernetes and Kafka in rapidly growing environments.
• Drive self-directed projects from concept to deployment with minimal oversight.
• Optimize infrastructure with Terraform and Azure for reliability and efficiency.
• Collaborate on safety research, translating insights into deployable engineering solutions.
• Monitor AI model performance, identifying alignment drifts and implementing fixes.
• Respond to production emergencies, restoring systems swiftly to minimize downtime.
• Contribute to OpenAI's safety tooling ecosystem, enabling safer AGI development.

These responsibilities position you at the intersection of software engineering, AI safety, and systems reliability, ensuring OpenAI's platforms remain trustworthy.

Qualifications

To excel, you should have:

• Proven experience in production services at high-growth companies.
• Expertise debugging and resolving live system issues rapidly.
• Background in content safety, fraud, abuse systems, or keen interest in AI safety.
• Strong programming in Python, C++, Rust, or Go.
• Nuanced understanding of AI capability-risk trade-offs.
• Ability to innovate risk mitigations without UX compromises.
• Pragmatic approach to engineering decisions.
• Excellent project management and self-direction.
• Experience with ML model deployment or eagerness to learn.
• Familiarity with our stack or quick learning ability.

Senior-level experience (5+ years) in scalable systems is preferred. We're seeking pragmatic builders excited by AI's real-world impact.

Salary & Benefits

Compensation for Software Engineer, AI Safety roles at OpenAI in San Francisco ranges from $220,000 to $380,000 USD yearly, including base salary, equity, and bonuses. Exact offers depend on experience and skills.

Benefits include comprehensive health coverage, 401(k) matching, unlimited PTO, professional development stipends, mental health support, parental leave, fitness reimbursements, commuter benefits, and more. Enjoy cutting-edge hardware, meal credits, and opportunities for global impact.

Why Join OpenAI?

OpenAI leads AI innovation with products like ChatGPT, pushing boundaries while prioritizing safety. Work with top talent on mission-critical safety systems that protect billions. Our San Francisco HQ fosters collaboration in a vibrant tech ecosystem. You'll gain exposure to state-of-the-art AI research, scale massive systems, and contribute to humanity's future. With a culture valuing diverse voices, pragmatic engineering, and bold impact, OpenAI offers unmatched growth for AI safety engineers.

Read more about our safety approach.

How to Apply

Ready to advance AI safety? Submit your resume, GitHub/portfolio, and a note on why you're passionate about OpenAI's mission. Highlight relevant projects in safety, scaling, or ML. We review applications on a rolling basis—apply now to join our Safety Systems team in San Francisco!