Job Description

Accountabilities:

• Conduct red-team testing of conversational AI models and agents to identify jailbreaks, prompt-injection vulnerabilities, misuse scenarios, and opportunities for bias exploitation.
• Develop creative and realistic adversarial prompts and attack scenarios designed to probe model weaknesses.
• Generate high-quality human evaluation data by annotating model failures, classifying vulnerabilities, and identifying systemic safety risks.
• Apply established taxonomies, benchmarks, testing frameworks, and playbooks to ensure consistent and rigorous evaluations.
• Document findings in a reproducible manner through detailed reports, datasets, attack cases, and supporting evidence.
• Clearly communicate technical findings and risk assessments to both technical and non-technical stakeholders.
• Adapt testing approaches across different projects, AI systems, use cases, and customer requirements while maintaining consistent quality standards.
• Contribute insights that help improve AI safety practices, model robustness, and risk mitigation strategies.

Requirements:

• Fluent or professional-level proficiency in both English and Norwegian, with strong written and verbal communication skills.
• Prior experience in AI red teaming, adversarial AI work, cybersecurity, penetration testing, or socio-technical risk assessment.
• Demonstrated ability to systematically probe complex systems for vulnerabilities, unexpected behaviors, and potential misuse.
• Experience applying structured frameworks, taxonomies, benchmarks, or testing methodologies to security or safety evaluations.
• Strong analytical and critical-thinking skills, combined with the creativity needed to develop novel probing strategies.
• Ability to clearly document technical findings and communicate them effectively to both technical and non-technical audiences.
• Strong attention to detail and commitment to producing reproducible, high-quality evaluation results.
• Adaptability and ability to move efficiently between different projects, models, use cases, and customer requirements.
• Experience in adversarial machine learning, cybersecurity, or socio-technical risk is preferred.
• Creative-probing skills developed through psychology, acting, creative writing, or related disciplines are an advantage.

Benefits:

• Competitive contract compensation of $48–$62 per hour
• Fully remote work environment.
• Flexible, project-based working structure.
• Opportunity to contribute directly to the safety and responsible development of advanced AI systems.
• Exposure to diverse AI models, agents, safety challenges, and adversarial testing methodologies.
• Opportunity to apply both technical and creative problem-solving skills to emerging AI safety challenges.
• Work across varied projects and use cases with the opportunity to expand expertise in AI safety and security.
• Meaningful contribution to improving the robustness, reliability, and responsible deployment of conversational AI.