AI Safety Practitioner
Pubblicato il 16-08-2026 - Mercor in Roma
We are seeking experienced AI Safety Practitioners to evaluate the safety, quality, and alignment of frontier AI models across complex, policy-sensitive, and ambiguous ("grey-area") topics. You will assess AI-generated responses, apply safety policies, and help improve model behavior through structured evaluations and feedback.
Se sta pensando di inviare una candidatura, si ricordi di premere il pulsante apposito dopo aver letto l'intera descrizione.
Responsibilities
-
- Evaluate AI-generated responses for safety, factual accuracy, policy compliance, and overall quality.
- Review content involving misinformation, political persuasion, self-harm, violence, cyber, biosecurity, and other sensitive domains.
- Apply and refine evaluation rubrics for RLHF, SFT, and AI safety benchmarking.
- Identify unsafe outputs, hallucinations, reasoning failures, and policy violations.
- Provide structured feedback to improve model alignment and safety performance.
- Collaborate with AI researchers and safety teams on ongoing evaluation initiatives.
Required Qualifications
-
- Bachelor's degree or higher in Journalism, Communications, Psychology, Sociology, Public Policy, Law, Biology, Chemistry, Computer Science, or a related discipline.
- 5+ years of professional experience in AI Safety, Trust & Safety, journalism, public policy, scientific research, security, or a related field.
- Excellent written English, critical thinking, and analytical reasoning skills.
- Ability to consistently evaluate nuanced and policy-sensitive scenarios.
Preferred Qualifications
-
- Experience with AI Safety, RLHF, SFT, Trust & Safety, or AI evaluation.
- Familiarity with safety policies, content moderation, or evaluation rubric development.
- Experience reviewing complex, high-risk, or ambiguous content.
Why Join?
-
- Shape the safety and behaviour of frontier AI models used by millions worldwide.
- Work on challenging, real-world safety evaluations across nuanced and high-impact domains. xbvzes