About the Team
The Safety Product (or Platform Responsibility) team is at the forefront of building and optimizing content safety systems. With a focus on optimising and advancing content safety, we leverage advanced large language models to enhance review efficiency, risk control, and user trust. Working closely with business and technical stakeholders, we deliver scalable solutions that keep pace with rapid global growth.
Responsibilities:
- Risk Research & Investigation: Conduct in-depth investigations into external incidents and signals identified across TikTok to understand how safety risks emerge, spread, and reach users. Identify recurring risk patterns and translate findings into clearly defined, testable research questions.
- Safety Evaluation Framework Development: Design scenario-based safety evaluation benchmarks grounded in real-world user experiences. Define user personas, behavioral journeys, environmental conditions, and expected safety outcomes to develop reproducible test cases and consistent, objective evaluation criteria.
- Adversarial Testing & Risk Assessment: Design and execute exploratory, adversarial, and cross-platform safety evaluations to assess TikTok's effectiveness in preventing harm and protecting users. Identify vulnerabilities, potential failure points in existing safeguards, and edge cases involving legitimate user behavior.
- Evaluation Methodology & Continuous Improvement: Continuously refine evaluation scenarios, test cases, and methodologies to keep pace with evolving risk patterns, evasion techniques, and user behaviors. Maintain standardized benchmarks alongside exploratory test cases to ensure evaluation coverage, comparability, and reliability.
- Cross-functional Collaboration & Risk Mitigation: Partner with Policy, Product, Engineering, Data, and Operations teams to translate research and evaluation findings into actionable risk mitigation recommendations. Drive cross-functional safety initiatives and conduct follow-up evaluations to measure the effectiveness of implemented improvements.
Qualifications
Minimum Qualifications
- Bachelor's degree or above, with at least 2 years of relevant professional experience. Backgrounds in risk identification, content safety and moderation, risk strategy, red teaming, model evaluation, investigative research, or journalism are welcome.
- Good investigative and analytical skills, with the ability to identify and diagnose complex problems, develop actionable solutions, prioritize improvements, and drive issue reproduction, implementation, and effectiveness measurement.
- Good interest in internet products and awareness of emerging AI technologies and industry trends. Familiarity with leading generative AI products, such as ChatGPT and Gemini, is expected; hands-on experience with AI model training or evaluation is a plus.
- Good analytical mindset and data literacy, with the ability to identify patterns, synthesize findings, generate actionable insights, and proactively drive problem resolution.
- Solid understanding of online community safety, with strong risk awareness and a sense of responsibility. Ability to proactively identify potential safety risks, content-related issues, and operational gaps, and develop practical mitigation strategies.
- Professional proficiency in written and spoken English, with the ability to communicate and collaborate effectively in an English-speaking work environment.
