- Company Name
- GoFundMe
- Job Title
- Adversarial AI Engineer
- Job Description
-
**Job title**
Adversarial AI Engineer
**Role Summary**
Design, execute, and automate adversarial testing of large language models (LLMs), agentic AI systems, recommendation algorithms, and fraud‑detection tools to strengthen AI safety and resilience. Lead red‑team operations, develop real‑time defense mechanisms, and shape governance, policy, and incident response for AI systems at a global scale.
**Expectations**
- 6–8 years cybersecurity experience with a strong focus on AI/ML security or adversarial machine learning.
- 2+ years specialized in LLM security (prompt injection, jailbreaking, adversarial prompt crafting).
- Proven red‑team/penetration‑testing background targeting AI systems.
- Deep understanding of ML fundamentals, neural networks, transformers (GPT, LLaMA, Claude, BERT).
- Proficiency in Python and ML frameworks (TensorFlow, PyTorch, Hugging Face).
- Experience with threat modeling, security architecture, and cloud controls (AWS, GCP, Azure).
**Key Responsibilities**
- Execute adversarial tests on LLMs, agentic AI, recommendation models, and fraud‑detection tools using prompt injection, jailbreaking, data poisoning, model extraction, and membership inference.
- Build synthetic attack datasets tailored to fundraising and trust & safety scenarios.
- Develop automated testing pipelines integrated into CI/CD, reusable robustness libraries, and real‑time detection for prompt injection and model evasion.
- Implement input validation, output filtering, adversarial training, differential privacy, and robustness certification.
- Partner with Trust & Safety, Product Security, and Data Science to mitigate algorithmic bias, enhance fraud countermeasures, and align with AI governance.
- Establish AI security policies, training, and deployment review processes aligned with NIST AI RMF; design monitoring and incident‑response workflows.
- Conduct research on emerging attack vectors, contribute to open‑source tools, and publish externally to position the organization as a leader in AI security.
**Required Skills**
- Python, TensorFlow, PyTorch, Hugging Face, LangGraph, AutoGen, CrewAI, Google ADK, Pydantic AI.
- Adversarial attack techniques: data poisoning, model evasion, membership inference, model extraction.
- Defense mechanisms: adversarial training, input sanitization, differential privacy, robustness certification.
- Knowledge of OWASP Top 10 for LLMs, MITRE ATLAS, NIST AI RMF.
- Threat modeling, security architecture, cloud security controls (AWS, GCP, Azure).
**Required Education & Certifications**
- Bachelor’s or Master’s degree in Computer Science, Cybersecurity, Data Science, or related field.
- Relevant certifications (CEH, OSCP, or equivalent) preferred; certifications in AI/ML security or cloud security (e.g., GCP Security Engineer, AWS Security Specialty) are a plus.
San francisco, United states
Hybrid
Junior
20-09-2025