Red Team Engineer Safeguards

About Anthropic

Anthropic’s mission is to create reliable, interpretable, and steerable AI systems. Our team of researchers, engineers, policy experts, and business leaders is rapidly growing to build safe, beneficial AI.

About the Role

The Safeguards team at Anthropic is seeking a Red Team Engineer Safeguards to uncover vulnerabilities in our AI systems using an adversarial approach. The role spans technical infrastructure vulnerabilities to emerging risks from advanced AI capabilities.

Responsibilities

• Conduct adversarial testing across product surfaces with creative attack scenarios.
• Research and implement novel testing for emerging capabilities.
• Design and execute full kill chain attacks simulating real-world threat actors.
• Build systematic testing methodologies and automated frameworks for continuous assessment.
• Collaborate with Product, Engineering, and Policy teams to drive improvements.
• Help establish metrics for measuring detection effectiveness of novel abuse.

Qualifications

• Experience in penetration testing, red teaming, or application security.
• Expertise with security testing tools such as Burp Suite, Metasploit, and custom scripting.
• Proven ability to discover novel attack vectors and chain vulnerabilities.
• Strong communication skills with ability to explain technical concepts.
• Bachelor's degree in a related field or equivalent experience.

Preferred Skills

• Experience with AI/ML security or adversarial machine learning.
• Familiarity with API security, rate limiting, and business logic vulnerabilities.
• Experience in anti-fraud, trust & safety, or abuse prevention systems.
• Understanding of distributed systems and advanced AI safety considerations.

Compensation & Logistics

Annual Salary: $300,000—$320,000 USD.

This role follows a location-based hybrid policy, expecting staff to be in-office at least 25% of the time. Visa sponsorship is available subject to eligibility.

How We're Different

At Anthropic, we focus on high-impact AI research with collaborative, big science efforts. We emphasize communication and a unified team approach to tackle significant AI safety challenges.

Key skills/competency

• penetration testing
• red teaming
• adversarial testing
• web security
• automation
• vulnerability assessment
• attack simulation
• scripting
• infrastructure security
• AI safety