Abstract: As generative AI advances, it introduces both unprecedented opportunities and complex threats within the cybersecurity ecosystem. This chapter examines how generative models can be weaponized to create synthetic attacks—deepfake spear-phishing, adversarial perturbations, automated malware, and identity spoofing—while also powering novel defensive strategies. It explores foundational concepts in adversarial machine learning, synthetic data generation for intrusion detection, AI-powered threat modeling, and automated vulnerability assessment. The role of generative models in privacy preservation—through federated learning, differential privacy, homomorphic encryption, and synthetic privacy-preserving datasets—is also analyzed. Real-world case studies illustrate evolving security challenges across finance, government, and critical infrastructure. The chapter concludes with a discussion of governance frameworks, digital trust mechanisms, and the future of resilient AI systems capable of defending against intelligent adversaries. Keywords: Cybersecurity; Adversarial AI; Deepfakes; Privacy Preservation; Digital Trust; Synthetic Attacks