As generative AI accelerates enterprise innovation, it introduces unprecedented security challenges that demand holistic, domain-specific frameworks. This paper proposes a comprehensive security architecture tailored to enterprise-scale generative AI deployments. The framework addresses five core pillars: infrastructure security, data protection, application security, responsible AI implementation, and regulatory compliance. Drawing from cloud-native principles, emerging AI governance standards, and real-world case studies, this paper outlines actionable strategies to mitigate risks such as prompt injection, data leakage, model manipulation, and compliance violations. It emphasizes the importance of integrated governance, ethical oversight, and secure-by-design architectures to enable sustainable, scalable, and compliant GenAI adoption. The framework supports security and innovation co-evolution, helping organizations unlock AI's full potential while protecting critical assets and maintaining trust.