Mercor50.5/hr
AI Red-Teamer — Adversarial AI Testing (Advanced); English & Korean
Test conversational AI systems with advanced adversarial techniques and document vulnerabilities.
AI Safety·Recent·📍 USA, South Korea
Apply Test conversational AI systems with advanced adversarial techniques and document vulnerabilities.
Red team AI models through structured adversarial methods and uncover vulnerabilities missed by automated testing.
Conduct adversarial testing of conversational AI models, identify systemic risks, and deliver reproducible red team artifacts.
Probe conversational AI systems using adversarial techniques such as jailbreaks, prompt injections, and bias exploitation to surface vulnerabilities and generate high-quality red team data.
Contribute clinical expertise across multiple medical specialties to healthcare-focused AI projects.
Provide financial expertise and reasoning support for AI-driven finance use cases.
Evaluate LLM behavior and validate large-scale code repositories.
Validate Docker-based workflows, CI/CD pipelines, and data processing systems.
Conduct advanced mathematical research and reasoning to support AI evaluation.
Provide advanced academic expertise across science and mathematics domains.
Analyze financial data and business processes to support decision-making.
Support real estate sales processes through client interaction.