AI & Cybersecurity Red-Teaming Lab

Initiated the design of an AI & Cybersecurity Red-Teaming Lab to investigate jailbreak vulnerabilities in large language models, iOS devices, and federated learning environments. Developed experimental frameworks for adversarial prompt injection, malicious client simulations, and Deepfake-based attacks on medical imaging systems. Proposed layered defenses, including guardrails for LLMs, forensic imaging verification, and federated anomaly detection, to build resilience against emerging threats.

AI and mobile device security through adversarial red-teaming, jailbreak analysis, and Deepfake attack simulation