Role Description : Your primary goal will be to identify and analyze potential safety policy violations, such as Hate Speech, Harassment, Sexually Explicit Content, Dangerous and Harmful Acts, and Child Safety concerns. The detailed feedback and insights you provide will be instrumental in training, fine-tuning, and improving the safety guardrails of our models.

Key Responsibilities

Multi-Modal Content Review: Systematically review and analyze AI-generated content (text, images, audio, video) to detect violations of our safety policies.
Policy Application and Annotation: Apply a complex set of safety policies with high accuracy and consistency. Provide detailed labels and annotations that capture the nuance of potential harms.
High-Quality Feedback Loop: Author detailed, data-driven feedback and rationale for your evaluations, which will be used directly by our engineering and research teams to improve model performance.
Trend Identification: Proactively identify, analyze, and document emerging trends, loopholes, and adversarial attack vectors (e.g., new forms of hate speech, jailbreaking prompts) that challenge our safety systems.
Policy Refinement: Contribute to the development and refinement of safety policies by providing expert feedback on policy clarity, edge cases, and operational feasibility.
Calibration and Quality Assurance: Participate in regular calibration sessions with team members to ensure consistent and fair application of policies.

Required Qualifications (Must-Haves)

Bachelor's degree or equivalent practical experience.
[7+] years of professional experience in Trust & Safety, content moderation, policy enforcement, or a related field.
Deep subject matter expertise in all of the following areas: Hate Speech, Harassment, Sexually Explicit Content, Child Safety, Self-Harm, Violence & Incitement, or Dangerous/Illegal Content.
Exceptional analytical skills and the ability to make consistent, high-judgment decisions in ambiguous situations.
Extremely high attention to detail and a commitment to quality.
Strong written and verbal communication skills, with an ability to articulate complex issues clearly and concisely.
Demonstrated resilience and emotional maturity to handle exposure to potentially disturbing or sensitive content on a regular basis.

Preferred Qualifications (Nice-to-Haves)

Proven experience working with safety evaluation across multiple modalities (e.g., experience with both text and image/video safety).
Familiarity with the landscape of Large Language Models (LLMs), diffusion models, or other generative AI systems.
Experience in a data-driven environment; ability to use data analysis tools (e.g., SQL, spreadsheets, data visualization tools) to derive insights is a plus.
Experience contributing to the development or iteration of content policies at a technology company.
Proficiency in one or more languages in addition to English, to help evaluate content from a global perspective.
A keen understanding of global and cultural nuances that affect content interpretation.

Content Safety Expert