In the rapidly expanding digital research landscape, curation is crucial for sustaining high-quality Open Educational Resources (OER) especially in research data management (RDM). As OER ecosystems grow, the proliferation of both human- and machine-generated content presents significant challenges for maintaining discoverability and educational relevance. The DALIA portal, a specialized search engine for OER on data literacy, is designed to ensure that all indexed resources meet high educational and technical standards across diverse scientific domains. This contribution examines how a comprehensive federated curation approach—drawing on distributed expertise from various scientific communities, National Research Data Infrastructure (NFDI) consortia, data competency centers, and federal RDM initiatives—will provide the foundation for DALIA's quality assurance framework. Unlike traditional centralized curation models, DALIA's federated approach will enable decentralized sourcing while ensuring semantic coherence and metadata integrity across platforms. This system will overcome the limitations of centralized quality control by leveraging domain-specific knowledge from multiple scientific communities. It will support customizable metadata schemas, such as the DALIA Interchange Format [1], which accommodate discipline-specific terminology, and community-specific collections. Additionally, collaborative curation tools will enable community-driven quality assessment. Through these mechanisms, scientific communities will actively participate in building and maintaining high-quality educational resource collections tailored to their specific needs. DALIA's comprehensive quality assurance system will integrate both AI-enhanced capabilities and API-driven content acquisition to create a sophisticated curation ecosystem. A key innovation will be the embedding of artificial intelligence in the curation interface to assist subject-matter experts in efficiently classifying, annotating, and enhancing resource metadata. This AI augmentation will include semantic tagging, content summarization, quality assessment prompts, and cross-referencing with existing taxonomies. Simultaneously, standardized APIs and persistent resource identifiers will automatically harvest metadata from source repositories, pre-populating curation forms with essential information. Comparative evaluation of leading OER platforms will reveal that without robust curation, even large-scale repositories fail to deliver relevant results for nuanced or interdisciplinary topics. In contrast, DALIA's approach will significantly improve search precision, allowing users to retrieve contextually aligned, pedagogically appropriate materials more reliably. By positioning this federated curation approach at the core of its infrastructure, DALIA will develop from a basic aggregator into an intelligent educational resource hub that actively supports research data management education. This work will highlight curation as a foundational layer of OER infrastructure—one that must evolve in tandem with the accelerating pace of automated content generation. It also advocates for the strategic integration of human-AI collaboration in curation workflows as a necessary response to the growing complexity and scale of educational resource management in the digital era. Currently, DALIA indexes a wide spectrum of data literacy tools, training materials, and best practices, and welcomes materials from all sides, nationally and also internationally. The platform is under continuous development with a strong focus on community engagement. Researchers, educators, data stewards, and professionals from libraries and archives are invited to contribute content, help shaping metadata standards, and provide feedback to ensure that DALIA stays responsive, inclusive, and aligned with the evolving data literacy landscape. By fostering collaboration and co-creation, DALIA strives to become a sustainable pillar of the academic data culture.