Security Engineer

Magic’s mission is to build safe AGI that accelerates humanity’s progress on the world’s most important problems. We believe the most promising path to safe AGI lies in automating research and code generation to improve models and solve alignment more reliably than humans can alone. Our approach combines frontier-scale pretraining, domain-specific RL, ultra-long context, and inference-time compute to achieve this goal.

About the role:

Increasingly capable AI systems are already finding vulnerabilities and carrying out attacks with less human involvement. As these capabilities advance, we need to build defenses that can anticipate and respond to sophisticated, autonomous attacks without relying on humans to recognize every threat or take action to stop it.

As a Security Engineer at Magic, you'll secure the systems we rely on today while building new software and defenses against increasingly sophisticated AI-driven attacks. You'll work directly with researchers and engineers to protect our infrastructure and build security into the way we develop and deploy our models. You will help us make the secure path the easy, productive path.

You'll also work closely with our Alignment team to identify and address security risks associated with increasingly capable AI systems. Security alone cannot protect us from the risks of misaligned superintelligence, and stronger defenses could mask those risks if progress on alignment doesn't keep pace. Your work will help ensure that Magic's models are a safe net-benefit to the world.

What you might work on: 

• Design, build, and maintain security tools and automation used by Magic's engineering teams

• Secure developer infrastructure, CI/CD pipelines, build systems, and production services without unnecessarily slowing down development

• Work with our sandboxing team to build secure isolation for untrusted code, agents, and model execution

• Build defenses against emerging AI security threats, including 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

• Investigate and respond to security incidents and active threats

What we’re looking for: 

• Software engineering skills with experience building production systems in Rust, Go, Python, C++, Kotlin, TypeScript, or similar languages

• Deep understanding of attacker tactics and techniques, including the MITRE ATT&CK framework

• Up-to-date knowledge of emerging AI security threats, real-world attacks, and the evolving capabilities of frontier AI systems

• Willingness to respond to critical security incidents whenever they arise

• Experience securing complex distributed systems and large-scale production infrastructure

• Ability to think like an attacker, identify exploitable weaknesses, and engineer effective defenses

• Strong expertise in Linux/UNIX, networking, authentication, authorization, and systems security

• Experience building security automation and AI agents that autonomously detect, investigate, and respond to threats

• Exceptional personal integrity and sound judgment, especially when operating independently and without oversight

Our culture

• Integrity. Words and actions should be aligned

• Hands-on. At Magic, everyone is building

• Teamwork. We move as one team, not N individuals

• Focus. Safely deploy AGI. Everything else is noise

• Quality. Magic should feel like magic

Magic strives to be the place where high-potential individuals can do their best work. We value quick learning and grit just as much as skill and experience.

Compensation, benefits, and perks (US):

• Annual salary ranges between $225K - $550K based on experience

• Equity is a significant part of total compensation, in addition to salary

• 401(k) plan with 6% salary matching

• Generous health, dental and vision insurance for you and your dependents

• Unlimited paid time off

• Visa sponsorship and relocation stipend to bring you to SF, if possible

• A small, fast-paced, highly focused team