Certified AI Red Team Engineer (CARTE) Training

Jacksonville professionals earn CARTE through the University of North Florida's Professional and Lifelong Learning partnership with Applied Technology Academy. This is the Certified AI Red Team Engineer program — a hands-on AI security certification, delivered live online or in person by ATA's practitioner instructors.

Coursework centers on exploiting, testing and securing real AI systems, from prompt injection to model exploitation, on intentionally vulnerable targets — the preparation Northeast Florida employers expect for practitioners who prove AI security skill through a flag-based practical exam. UNF PLL students get ATA's hands-on labs, unlimited mentoring and 5-star student support from first class to certification.

The UNF Professional and Lifelong Learning team handles enrollment — pick a session below or reach out and we'll map your schedule, funding and prep together.

Course Outline

• Module 1: Foundations of AI Security
  • Understanding AI systems, threat landscape, MITRE ATLAS, OWASP Top 10 for LLMs.
• Module 2: AI System Internals
  • Transformers, embeddings, RAG systems, agents, and how they work under the hood.
• Module 3: Guardrails and Safety Mechanisms
  • Input filters, system prompts, RLHF, Constitutional AI, output filters, and bypassing defenses.
• Module 4: Prompt Injection
  • Basic to advanced injection techniques, system prompt extraction, and defense bypassing.
• Module 5: Jailbreaking
  • Role-playing attacks, DAN techniques, multi-step jailbreaks, and adversarial prompts.
• Module 6: Data Poisoning
  • Backdoor attacks, training data contamination, and manipulating AI behavior.
• Module 7: RAG Exploitation
  • Document injection, retrieval manipulation, and vector database attacks.
• Module 8: Agent Red Teaming
  • Tool injection, action hijacking, and compromising autonomous AI agents.
• Module 9: Model Inversion
  • Extracting training data, membership inference, and model stealing techniques.
• Module 10: Adversarial ML
  • Adversarial examples, evasion attacks, and fooling neural networks.
• Module 11: API Security
  • Rate limiting bypass, token manipulation, and exploiting AI API endpoints.
• Module 12: Real CVEs
  • Exploit actual CVEs in production AI systems. Hands-on with real vulnerabilities.
• Module 13: Professional Red Teaming
  • Methodology, ethics, and running complete red team assessments.
• Module 14: Ethics and Responsible Disclosure
  • Legal considerations, ethical frameworks, responsible disclosure practices, and professional conduct.
• Module 15: Documentation & Evidence
  • Documenting findings, capturing evidence, and organising results from AI security assessments.
• Module 16: Pitching AI Red Team Engagements
  • Sales strategies, proposal writing, pricing models, and winning AI security consulting contracts.
• Module 17: Exam Preparation
  • Comprehensive review, exam strategy, time management, and final preparation for the practical exam.
• Module 18: Final Practical Exam
  • 7-day hands-on assessment. Hack 15 real vulnerable AI systems and capture 30 flags. 80% pass mark required.