B. Bhanu Sree 01 01 The rapid adoption of cloudnative applications microservices hybrid cloud platforms mobile systems Internet of Things environments artificial intelligence services and distributed enterprise applications has significantly increased organizational dependence on application programming interfaces APIs provide standardized mechanisms for communication among internal applications external partners cloud services data platforms machine learning systems and customerfacing applications However conventional perimeterbased security models are increasingly inadequate because modern enterprise services operate across dynamic network boundaries multiple identity domains heterogeneous infrastructure and continuously changing workloads Attackers can exploit compromised credentials excessive privileges insecure tokens weak authorization vulnerable endpoints service impersonation lateral movement API abuse data leakage and communication manipulation This paper proposes a ZeroTrust API Security Architecture for protecting distributed enterprise services The proposed framework applies the principle of never trust implicitly and continuously verify every request identity device workload service token and access context The architecture integrates identitycentric authentication workload identity mutual transport authentication token validation attributebased authorization leastprivilege access microsegmentation API gateways service mesh controls secrets management behavioral analytics risk scoring continuous verification policy enforcement anomaly detection secure API lifecycle management and automated incident response A layered methodology is introduced to discover APIs classify service sensitivity establish trusted identities validate contextual attributes evaluate request risk enforce dynamic access decisions inspect communication behavior detect anomalies and continuously update security posture Machine learning models including Random Forest Gradient Boosting Artificial Neural Networks and temporal analytics are incorporated for API abuse detection and adaptive risk assessment A representative prototypeoriented evaluation indicates that the proposed architecture can improve unauthorized request detection reduce lateral movement opportunities strengthen credential protection decrease excessive access and provide faster response to anomalous API behavior compared with static perimeteroriented security The architecture also incorporates productionscale machine learning governance secure service integration cyberrisk modeling cloudnative sustainability and enterprise observability The study demonstrates that Zero Trust provides a scalable security foundation for protecting distributed enterprise API ecosystems where network location alone cannot be treated as evidence of trust Copyright (c) Journal of Science Engineering Technology and Management Science. All rights reserved 2026