Agentic AI, characterized by its ability to autonomously code, deploy applications, and integrate with production environments, necessitates a shift from traditional security models to a Zero-Trust approach, which emphasizes continuous verification, minimized permissions, and adaptive guardrails. This paradigm shift addresses the unique risks posed by agentic AI, such as secret leakage and autonomous misconfigurations, by embedding security measures directly into AI workflows and decision-making processes. The Zero-Trust model for agentic AI requires continuous identity verification, real-time monitoring, and adaptive responses to ensure that AI-generated code aligns with enterprise security standards. Furthermore, it involves automation-first security practices, cloud-native controls, and predictive analytics to preemptively thwart potential threats and maintain compliance with regulatory standards. As agentic AI continues to grow, projected to expand from USD 5.25 billion in 2024 to over USD 100 billion by 2034, organizations must integrate these security frameworks to harness AI's capabilities while safeguarding against new vulnerabilities.