Cybersecurity Software Engineer – AI Safety
About the Role
A leading AI safety initiative is seeking cybersecurity-focused software engineers to contribute technical expertise toward the evaluation and improvement of advanced AI systems. This opportunity centers on strengthening model performance, safety, and reliability when handling cybersecurity-related topics.
This opportunity is ideal for professionals with strong cybersecurity and software engineering backgrounds who can assess technical content, apply structured judgment, and communicate complex concepts with clarity. Prior experience in artificial intelligence or machine learning is not required.
The work involves creating cybersecurity-focused evaluation scenarios, reviewing AI-generated responses, classifying interactions according to established guidelines, and providing expert feedback to improve model behavior. Success in this role requires strong technical reasoning, sound security judgment, and attention to detail.
What You'll Do
- Develop expert-level prompts covering a range of cybersecurity topics
- Evaluate AI-generated responses for technical accuracy, relevance, and safety
- Annotate and classify prompts, conversations, and model outputs using structured evaluation frameworks
- Identify weaknesses, risks, and areas for improvement in model performance
- Apply cybersecurity expertise to assess the handling of sensitive or dual-use information
- Provide detailed written feedback to improve AI model quality and reliability
- Review technical content for consistency with cybersecurity best practices
- Collaborate within project workflows to maintain evaluation quality standards
- Support ongoing AI safety and model assessment initiatives
Requirements
- Bachelor’s or Master’s degree in Computer Science or a closely related field, or equivalent professional experience
- Minimum 5 years of software engineering experience in a professional environment preferred
- Strong understanding of cybersecurity principles, threats, vulnerabilities, and modern software systems
- Excellent technical reasoning and analytical problem-solving skills
- Strong written communication skills in English
- Ability to evaluate technical content with accuracy and consistency
- Sound judgment regarding responsible handling of security-related and dual-use information
- Ability to follow structured guidelines and evaluation frameworks
- Strong attention to detail and commitment to high-quality deliverables
- Ability to work independently in a fully remote environment
- Must be authorized to work as an independent contractor within the United States
- Availability for approximately 15–25 hours per week, with flexibility for additional hours when required
- Background in offensive security, penetration testing, vulnerability research, or related disciplines preferred
- Experience reviewing, grading, auditing, or red-teaming technical content preferred
- Experience with AI evaluation workflows, model testing, or tools such as ChatGPT preferred