The AI Safety Expert role at Mercor is focused on enhancing the safety of conversational AI systems through red teaming methodologies. The successful candidate will engage in various activities such as performing jailbreaks and prompt injections to identify vulnerabilities, generating human data, and documenting findings for customer use. Applicants should be fluent in English and Malayalam and possess prior experience in cybersecurity or adversarial work.