How to apply for Software Engineer, Infrastructure Security

About OpenAI

OpenAI is a frontier AI research and product company known for developing GPT-4, DALL·E, and ChatGPT, with a mission to ensure that artificial general intelligence benefits all of humanity. Working here means contributing to cutting-edge AI while tackling some of the most challenging security problems in high-performance computing. The company's focus on safety and alignment extends to its infrastructure, making this role critical to protecting the systems that train and serve advanced AI models.

About the role

As a Software Engineer on the Infrastructure Security team, you will design and build security services for OpenAI's supercomputing environment, which powers large-scale AI training and inference. You'll work at the intersection of hardware, operating systems, and networks to architect production-grade security solutions, automate risk detection, and embed security into high-performance compute clusters. This role is impactful because you'll directly protect the infrastructure that enables OpenAI's mission, ensuring that security is foundational to every system launch.

A typical day

A typical day might involve collaborating with infrastructure engineers to threat model a new cluster design, writing code to automate detection of anomalous network activity, and reviewing security architecture for a service that will handle sensitive model data. You'll balance proactive security design with reactive risk mitigation, ensuring that OpenAI's supercomputing environment remains secure without slowing down innovation.

Who OpenAI is looking for

• Has hands-on experience designing and building security services for large-scale, high-performance computing environments (e.g., GPU clusters, supercomputers, or large cloud deployments).
• Possesses deep expertise in security architecture across hardware, operating systems, and networks, with a track record of securing both cloud and on-prem infrastructure.
• Is skilled at developing automation and detection tooling to identify and mitigate security risks, ideally using languages like Python, Go, or Rust.
• Has experience leading design reviews and threat modeling sessions, and can collaborate effectively with infrastructure and research engineers to integrate security into new systems.

Tips for this application

• Highlight any experience you have with securing high-performance computing (HPC) or GPU-accelerated environments, as OpenAI's supercomputing clusters are central to this role.
• Emphasize your ability to build automation and detection tools—provide specific examples of tools you've developed and the risks they mitigated.
• Demonstrate familiarity with OpenAI's mission and the unique security challenges of a frontier AI lab; mention how you've handled sensitive data or systems in previous roles.
• Showcase your experience with threat modeling and design reviews by describing a time you influenced the security architecture of a major system from the ground up.
• If you have contributed to open-source security projects or published research on infrastructure security, include links or references—OpenAI values technical depth and community impact.

What to cover in your cover letter

["Your specific experience securing large-scale compute environments, particularly if you've worked with GPUs or HPC clusters, and how that translates to OpenAI's supercomputing needs.", "Concrete examples of automation and detection tooling you've built, including the technologies used and the impact on risk reduction.", 'Your approach to collaborating with engineers to embed security into fast-moving development cycles, emphasizing your ability to balance security with performance and innovation.', "Why you are motivated to work at OpenAI specifically, and how you align with the company's mission of ensuring AGI benefits humanity, especially in the context of infrastructure security."]

Research before applying

• Read OpenAI's research papers and blog posts on infrastructure and security, such as those related to large-scale training and deployment, to understand the technical challenges they face.
• Learn about the 80,000 Hours career review on working at an AI lab to understand the ethical considerations and ensure your motivations align with the role.
• Investigate the specific security challenges of supercomputing environments, including GPU security, high-speed networking, and orchestration tools like Kubernetes.
• Familiarize yourself with OpenAI's products and mission, including recent developments in AI safety and how infrastructure security supports them.

Likely interview topics

Based on the job description, expect questions about:

• How would you design a security service to protect a large-scale GPU cluster used for training AI models? Walk through your architecture and threat model.
• Describe a time you automated detection of a security risk in a cloud or on-prem environment. What tools did you use, and how did you measure success?
• What are the unique security challenges of a hybrid cloud/on-prem environment, and how would you address them in a high-performance computing context?
• How do you approach threat modeling for a new system that needs to launch quickly? Give an example of a design review you led.
• OpenAI's infrastructure handles sensitive model weights and data. How would you ensure the integrity and confidentiality of these assets across hardware, OS, and network layers?

Common mistakes to avoid

• Focusing only on generic security experience without addressing the unique demands of high-performance computing or AI infrastructure.
• Underestimating the importance of automation and detection tooling—candidates who only talk about manual security reviews may not stand out.
• Ignoring OpenAI's mission and the ethical context of working at a frontier AI company; they want candidates who are thoughtful about the impact of their work.