The Zero-trust security architecture is a paradigm shift toward resilient cyber warfare. Although Intrusion Detection Systems (IDS) have been widely adopted within military operations to detect malicious traffic and ensure instant remediation against attacks, this paper proposed an explainable adversarial mitigation approach specifically designed for zero-trust cyber warfare scenarios. It aims to provide a transparent and robust defense mechanism against adversarial attacks, enabling effective protection and accountability for increased resilience against attacks. The simulation results show the balance of security and trust within the proposed parameter protection model achieving a high F1-score of 94\%, a least test loss of 0.264, and an adequate detection time of 0.34s during the prediction of attack types.