AI Safety Expert for English and Norwegian Red Teaming

Probe conversational AI systems for safety and misuse vulnerabilities in both English and Norwegian, producing high-quality adversarial test cases, annotations, and reproducible reports that help customers reduce bias, misinformation, and harmful behaviors. The work is text-based and focused on uncovering failures automated tests miss.

• Red team conversational AI models and agents, including jailbreaks, prompt injections, misuse cases, bias exploitation, and multi-turn manipulation
• Review AI outputs that touch on sensitive topics such as bias, misinformation, or harmful behaviors
• Generate human-labeled data: annotate failures, classify vulnerabilities, and flag systemic risks
• Follow established taxonomies, benchmarks, and playbooks to keep testing consistent and reproducible
• Document findings as actionable artifacts, including reports, datasets, and reproducible attack cases customers can act on

• Native fluency in English and Norwegian is required
• Prior red teaming experience, such as AI adversarial work, cybersecurity, or socio-technical probing
• Comfort with structured testing methods, using frameworks or benchmarks rather than ad hoc approaches
• Strong communication skills, able to explain risks to technical and non-technical stakeholders
• Ability to adapt and move across projects and customer contexts

• Adversarial machine learning: jailbreak datasets, prompt injection, RLHF or DPO attacks, model extraction
• Cybersecurity: penetration testing, exploit development, reverse engineering
• Socio-technical risk analysis: harassment or disinformation probing, abuse analysis, conversational AI testing
• Creative probing skills: psychology, acting, or writing for unconventional adversarial thinking

• Discovering vulnerabilities automated tests miss
• Delivering reproducible artifacts that customers can use to strengthen their AI systems
• Expanding evaluation coverage so fewer surprise failure modes appear in production
• Increasing customer trust in the safety of their AI systems because adversarial scenarios have already been exercised

• Remote engagement
• Hourly employment
• All tasks are text-based
• Participation in higher-sensitivity projects is optional, with clear topic briefings and wellness resources provided; topics will be communicated before exposure

48 - 62 hourly

• Native fluency in both English and Norwegian is required
• Ability to work remotely under an hourly engagement model