Member of Technical Staff, Defensive Security Engineer

Magic•San Francisco, CA

5d•$225,000 - $550,000•Remote

About The Position

As the first dedicated Defensive Security Engineer at Magic, you will be responsible for building the software, automation, and infrastructure to protect our models, systems, and engineering organization from both traditional and frontier AI cyber threats. This role involves responding to security threats and establishing best practices for security incident and threat response. You will operate at the intersection of software engineering, infrastructure, and security, collaborating with researchers, infrastructure engineers, and platform teams to minimize risk while maintaining engineering productivity. The goal is to make the secure path the easy and productive one. The role acknowledges the evolving threat landscape, including supply chain attacks, prompt injections, and other LLM-native attack vectors. As AI systems advance, security extends beyond traditional infrastructure defense to address autonomous vulnerability discovery, exploit chaining, prompt injection manipulation, and containment breaches. A strong emphasis is placed on AI safety, requiring collaboration with the alignment team to ensure Magic's models provide a net benefit to the world.

Requirements

Software engineering skills with experience building production systems in Rust, Go, Python, C++, Kotlin, TypeScript, or similar languages
A very good understanding of modern MITRE att&ck techniques
Actively engaged in the role of LLMs in securing and attacking software systems
On-call readiness 24/7, assisted by our team
Experience securing distributed systems, cloud infrastructure, or large-scale production environments
Ability to develop high-complexity cloud Linux-based exploits
Deep understanding of UNIX or other operating systems, networking, authentication, authorization, and modern security architecture
Experience building automation and tooling to secure fast-paced organizations
Track record of exceptional personal integrity, accountability and trustworthiness in high autonomy environments

Responsibilities

Design, build, and maintain security tooling and automation used by Magic's engineering teams
Secure developer infrastructure, CI/CD pipelines, build systems, and production services, ensuring platform tools are user-friendly
Support the sandboxes team in developing secure sandboxing and isolation for untrusted code, agent, and model execution
Build guardrails against emerging AI attack vectors such as prompt injection, tool abuse, exploit chaining, and autonomous model behavior
Protect model weights, research artifacts, datasets, and production infrastructure from unauthorized access, lateral movement, and exfiltration
Respond to security incidents and live threats