Mercor is building a remote red team for evaluating AI safety by probing conversational models with adversarial inputs. The role focuses on identifying vulnerabilities, annotating failures, and producing reproducible artifacts to strengthen client AI systems.
You’ll work across projects with transparent guidelines, extended to sensitive topics, with wellness resources and optional participation in higher-sensitivity work. English and Assamese required.
J-18808-Ljbffr
📌 Remote Ai Safety Red Team Specialist Perth
🏢 Mercor
📍 Perth