AI Safety Evaluation & Governance Product Manager Intern (TikTok-Platform Responsibility-Feed Safety) - 2027 Summer
Core
Design and execute evaluation plans for AI safety models, build high-quality ground truth datasets, and analyze model performance to ensure content safety decisions meet governance standards.
Role type
Intern product manager (AI safety evaluation & governance)
Builds
AI-powered safety systems for content review and risk governance at scale
Domain
Social media platform content safety, AI/ML governance, and trust & safety policy
Deliverable
production ML models
Required skills
applied statistics (sampling, hypothesis testing, confidence intervals), AI/ML concepts (classification, NLP, LLMs), structured problem decomposition, cross-functional communication, data sampling strategies, data cleaning pipelines, statistical error analysis
Preferred skills
prompt engineering, LLM-based evaluation, human-in-the-loop systems, experimental design, computational social science
Technologies
LLMs, agentic workflows
Responsibilities
Design evaluation objectives and metrics for AI safety models; build and maintain ground truth datasets with consistent labeling; analyze model performance using statistical methods; collaborate with engineers on model improvement and prompt design; communicate evaluation results and governance standards to stakeholders; improve evaluation processes and methodologies
Seniority
Intern
