As enterprise networks grow in scale and complexity, organisations struggle to maintain secure, consistent, and auditable configuration across distributed infrastructure. Manual configuration processes, disparate tools, and inconsistent permissions models introduce significant operational risk—creating opportunities for misconfiguration, outages, and security breaches. Trusted Agentic AI for Permissions Management addresses this challenge by combining secure automation with intelligent analysis to transform how corporate network configuration is managed. The project introduces an AI-enabled agent capable of understanding, validating, and enforcing network permissions and configuration policies across multi-vendor, multi-domain environments. Using a trusted, auditable, and explainable agentic AI framework, the solution can assist enterprise operations teams by: Analysing existing network configurations to identify risks, policy gaps, and inconsistencies. Recommending or applying configuration updates within defined permissions boundaries. Automating compliance checks and producing clear, human-readable explanations of actions taken. Reducing operational overhead while increasing accuracy and security across the network lifecycle. This Catalyst aims to demonstrate how next-generation, agent-driven automation can strengthen corporate defence postures, improve operational efficiency, and provide enterprises with a reliable foundation for secure connectivity. By showcasing real-world use cases and cross-industry collaboration, the project will contribute repeatable assets and best practices for bringing trusted agentic AI into mainstream network operations.