**Abstract:** Low-resource languages face an urgent threat of extinction, largely due to limited documented linguistic resources. This research proposes a novel Automated Cross-Lingual Semantic Alignment (ACSA) system utilizing Hierarchical Generative Adversarial Networks (HGANs) and a parallel morphology-aware embedding approach to dramatically accelerate the process of bilingual dictionary creation and semantic resource building for endangered languages. The system leverages readily available (though limited) parallel corpora and unsupervised morphology extraction to achieve a 4x improvement in alignment accuracy compared to traditional word alignment techniques and demonstrably facilitates the human linguist's ability to rapidly construct annotated resources. Our approach has the immediate potential to reverse language endangerment trends by promoting accessible and scalable documentation tools.