AI Content Red Team Analyst - Trust and Safety
Core
Conduct unstructured adversarial testing on TikTok's generative AI products and models to uncover emerging risks, failure modes, and abuse patterns.
Role type
Senior IC AI Content Red Team Analyst (Trust & Safety)
Builds
Safety testing playbooks, taxonomies, and internal knowledge bases for GenAI products
Domain
Generative AI safety and adversarial testing
Deliverable
production ML models
Required skills
adversarial testing, jailbreak analysis, LLM evaluation, risk assessment, prompt engineering, cross-functional collaboration, technical writing
Preferred skills
agentic AI tools, process automation
Technologies
LLMs, GenAI models
Responsibilities
Conduct structured adversarial testing on AI models, features, and policies; explore product behavior across contexts to identify model failure modes; investigate jailbreaks, evasions, and prompt-based attacks; document findings with risk descriptions and mitigation recommendations; partner with stakeholders to ensure mitigation validation; support development of testing playbooks and taxonomies; stay updated on emerging adversarial trends.
Seniority
Senior, hands-on IC