OpenAI· Safety Systems· San Francisco
Software Engineer, AI Safety
Comp$185K – $325K
Classified Tasks (21)
Automate 0%Augment 81%Human-Only 19%
Augment (17)
AI assists, human decides
Architect anti-abuse and content moderation infrastructure to protect the company and end users from unwanted behavior.
technical
Build anti-abuse and content moderation infrastructure to protect the company and end users from unwanted behavior.
technical
Maintain anti-abuse and content moderation infrastructure to ensure ongoing protection and reliability.
operational
Design systems that detect and prevent abuse across the platform.
technical
Implement systems that detect and prevent abuse across the platform.
technical
Design and implement systems that promote user safety and reduce platform risk.
technical
Collaborate with engineers and researchers to utilize industry-standard and novel AI techniques to measure AI models’ alignment to human values.
analytical
Monitor AI models’ alignment to human values using industry-standard and novel AI techniques.
analytical
Improve AI models’ alignment to human values based on measurement and monitoring results.
technical
Diagnose active incidents on the platform.
operational
Remediate active incidents on the platform to restore safe operation.
operational
Build tooling and infrastructure that address root causes of system failures.
technical
Design and implement trust and safety capabilities across the platform.
technical
Build and run production services in a high-growth, rapidly scaling environment.
operational
Debug live issues and restore systems quickly.
operational
Work on content safety, fraud, and abuse mitigation efforts.
technical
Deploy classifiers or machine learning models and work with modern ML infrastructure.
technical
Human-Only (4)
Requires human judgment
Assess trade-offs of capabilities and risks and navigate them to deploy products and features safely.
leadership
Critically assess risks of new products or features and devise solutions to mitigate those risks without harming the product experience.
leadership
Determine when to implement quick fixes versus invest in robust, lasting solutions.
leadership
Remove roadblocks and drive projects to completion with minimal guidance.
leadership
Job description
Software Engineer, AI Safety | OpenAI Careers ## Software Engineer, AI Safety Safety Systems - San Francisco Apply now(opens in a new window) **About the Team** The Safety Systems team is dedicated to ensuring the safety, robustness, and reliability of AI models and their deployment in the real world. Learn more about OpenAI’s approach to safety. Building on the many years of our practical alignment work and applied safety efforts, Safety Systems addresses emerging safety issues and develops new fundamental solutions to enable the safe deployment of our most advanced models and future AGI, to make AI that is beneficial and trustworthy. **About the Role** At OpenAI, we're dedicated to advancing artificial intelligence, and we know that creating a secure and reliable platform is vital to our mission. That's why we're seeking a software engineer to help us build out our trust and safety capabilities. In this role, you'll work with our entire engineering team to design and implement systems that detect and prevent abuse, promote user safety, and reduce risk across our platform. You'll be at the forefront of our efforts to ensure that the immense potential of AI is harnessed in a responsible and sustainable manner. **In this role, you will:** * Architect, build, and maintain anti-abuse and content moderation infrastructure designed to protect us and end users from unwanted behavior. * Work closely with our other engineers and researchers to utilize both industry standard and novel AI techniques to measure, monitor and improve AI models’ alignment to human values. . * Diagnose and remediate active incidents on the platform and build new tooling and infrastructure that address the root causes of system failure. **You might thrive in this role if:** * You have built and run production services in a high growth, rapidly scaling environment. * You can debug live issues and restore systems quickly. * You have worked on content safety, fraud, or abuse, or are motivated and excited to work on present-day (“now-term”) AI safety. * You have experience with Python or with modern languages such as C++, Rust, or Go, and are able to quickly ramp up on Python. * You understand the trade-offs of capabilities and risks and navigate them to deploy novel products and features safely. * You can critically assess risks of a new product or feature and devise innovative solutions to mitigate these risks without harming the product experience. * You’re pragmatic. You know when to build a quick, good-enough fix, and when to invest in a robust, lasting solution. * You possess strong project management skills. You are self-directed and can remove roadblocks to drive projects to completion with minimal guidance. * You’ve deployed classifiers or machine learning models, or are excited to learn about modern ML infra. **Our tech stack** * Our infrastructure is built on Terraform, Kubernetes, Azure, Python, Postgres, and Kafka. While we value experience with these technologies, we are primarily looking for engineers with strong technical skills who understand the fundamental problems these tools solve, and can quickly pick up new tools and frameworks. **About OpenAI** OpenAI is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence benefits all of humanity. We push the boundaries of the capabilities of AI systems and seek to safely deploy them to the world through our products. AI is an extremely powerful tool that must be created with safety and human needs at its core, and to achieve our mission, we must encompass and value the many different perspectives, voices, and experiences that form the full spectrum of humanity. We are an equal opportunity employer, and we do not discriminate on the basis of race, religion, color, national origin, sex, sexual orientation, age, veteran status, disability, genetic information, or other applicable legally p