Mercor
$60–70/hr
44 people hired
You will evaluate frontier AI models for safety, quality, factual accuracy, and policy alignment across complex and sensitive topics. | The role involves reviewing AI-generated responses on misinformation, politics, self-harm, violence, cybersecurity, biosecurity, and other high-risk domains. | You will apply structured safety policies and evaluation rubrics to identify unsafe outputs, hallucinations, and reasoning failures. | Your feedback will help improve model alignment and safety performance.
AI Safety Evaluation Expert