Senior Privacy Security Engineer (AI Red Teaming) - PDPO (Singapore)
Core
Conduct privacy security assessments and adversarial red teaming for Large Language Models (LLMs) and AI Agents to identify and mitigate risks in training data, model outputs, and inference services.
Role type
Senior IC privacy security engineer (AI red teaming)
Builds
Systematic and scalable model privacy evaluation frameworks, automated red teaming agents, and threat models for LLM Agent systems.
Domain
AI/ML security, LLM privacy, adversarial testing
Deliverable
production ML models
Required skills
LLM and Agent architecture understanding, prompt engineering, RAG, function calling, threat modeling, adversarial testing, membership inference analysis, automated agent development, privacy impact assessment
Preferred skills
CTF/HW competition experience, academic publications, vulnerability discovery via SRCs, penetration testing in production
Technologies
LangChain, AutoGen, LLM frameworks, cloud service architectures, database storage systems
Responsibilities
Evaluate privacy risks across model training data, outputs, and user data pipelines; develop threat models and conduct adversarial testing for prompt injection and data leakage; design automated red teaming agents; research emerging AI privacy challenges like model memorization and unlearning.
Seniority
Senior, hands-on IC