AI Red Teamer, Frontier AI Safety

Core responsibilities:

• Create adversarial prompts to test and improve the safety of frontier AI models
• Write creative scenarios designed to challenge AI systems' boundaries and safety measures
• Develop conversation flows that attempt to elicit unsafe or unintended AI responses
• Contribute to AI safety infrastructure by identifying potential vulnerabilities in AI models

This role helps tackle critical AI safety challenges by identifying vulnerabilities in frontier AI models, contributing to the development of safer AI systems that benefit society.

Required skills and qualifications:

• Strong storytelling and language skills
• Ability to think creatively about AI system vulnerabilities
• Understanding of AI safety challenges and frontier AI models
• Experience with adversarial testing or red teaming approaches

No specific benefits or compensation information provided in the job description.

July AI appears to be an organization focused on AI safety, specifically working on frontier AI models and their safety challenges.