Red Team Engineer, Safeguards
$320k - $405k • San Francisco, CA
Posted 8d ago
Job Location
San Francisco, CA
Tech Stack
Remote Work Policy
On-site
Categories
AI Engineer
About the job
Anthropic's Safeguards team is looking for a Red Team Engineer to ensure the safety of their AI systems and products. This role involves taking an adversarial approach to find vulnerabilities before malicious actors can exploit them. The work covers technical infrastructure weaknesses and emergent risks from advanced AI capabilities, focusing on broader safety implications and novel abuses unique to AI systems. You will simulate sophisticated threat actors, investigate potential abuse like account manipulation and payment fraud, and chain multiple attack vectors to achieve objectives.
Responsibilities
- Conduct adversarial testing across product surfaces, developing creative attack scenarios.
- Research and implement novel testing approaches for emerging capabilities like agent systems and tool use.
- Design and execute 'full kill chain' attacks emulating real-world threat actors.
- Build and maintain systematic testing methodologies for all system aspects.
- Develop automated testing frameworks for continuous assessment at scale.
- Collaborate with Product, Engineering, and Policy teams to implement improvements.
- Help establish metrics for measuring detection effectiveness of novel abuse.
Requirements
- Experience in penetration testing, red teaming, or application security.
- Experience in model jailbreaking and testing large-scale agentic workflows for prompt injection.
- Strong technical skills in web application security and security testing tools.
- Experience building custom automation, including LLM-specific testing frameworks.
- Proven track record of discovering novel attack vectors and chaining vulnerabilities.
- Public body of work (CVEs, blog posts, bug bounty reports).
- Strong written and verbal communication skills.
Benefits
- Annual compensation range: $320,000 - $405,000 USD
- Visa sponsorship available