The RAIDO project recently marked a significant milestone with its active participation at the 2025 IEEE International Conference on Cyber Humanities (IEEE-CH) in Florence, Italy. The project contributed substantially to the core program and the specialized RAIDO-GTAI workshop on “Green Trustworthy AI for Digital Humanities and Cultural Heritage,” demonstrating its commitment to sustainable and legally sound AI solutions across Europe.
Research Highlights in IEEE Xplore
Three key research papers stemming from the RAIDO project were presented and are now published in the IEEE Xplore IEEECH2025 conference proceedings:
- Energy-Efficient and Privacy-Preserving Federated Continual Learning for Cultural Heritage Preservation and Digital Humanities This paper introduces a Federated Continual Learning (FCL) framework that allows cultural institutions to train AI models (such as artifact classifiers) collaboratively without sharing sensitive raw data. By keeping data localized and only sharing model updates, the system ensures high privacy standards and significantly reduces the energy footprint of AI training, aligning with “Green AI” goals
- IEEE Xplore: https://ieeexplore.ieee.org/document/11279476/
- Zenodo: https://zenodo.org/records/18504685
- From Fragmentation to Cohesion: Multilingual Legal AI for Europe’s Evolving Data Landscape This research presents LexAid-EU, a jurisdiction-aware framework designed to solve “hallucinations” in general-purpose AI when interpreting complex laws like the GDPR and the AI Act. By utilizing Retrieval-Augmented Generation (RAG), LexAid-EU grounds its multilingual outputs in authoritative legal sources, ensuring transparency and compliance across European jurisdictions.
- IEEE Xplore: https://ieeexplore.ieee.org/document/11279282
- Zenodo: https://zenodo.org/records/18505056
- MemoryINP: Memory Enhanced Neural Processes for Understanding-Based Informed Meta-Learning To address the challenge of training AI with limited data, this paper introduces MemoryINP, which integrates a “differentiable memory” mechanism into Neural Processes. This enhancement allows models to effectively share and recall knowledge across tasks, making it ideal for the digital humanities where high-quality datasets are often scarce.
- IEEE Xplore: https://ieeexplore.ieee.org/document/11279273/
- Zenodo: https://zenodo.org/records/18504862
Moving Forward
Through these contributions, RAIDO has successfully demonstrated practical applications of its core pillars: Sustainability, Privacy, and Legal Compliance. The project’s support for the GTAI workshop fostered a collaborative environment for researchers and industrial partners (including MetaMind Innovations, SID, and others) to exchange insights on technology transfer and critical digital infrastructures.
The RAIDO-GTAI workshop will return to explore new frontiers in trustworthy AI. We invite researchers to submit their latest work according to the following timeline in the official IEEECH2026 page and the GTAI Workshop:
- Submission Deadline: April 12, 2026
- Notification of Acceptance: May 2, 2026
- Camera-Ready Submission: May 15, 2026
Find us on Social Media
