Semantics and Compliance at the Core of Trustworthy European Data Spaces: The DICE Alliance at EBDVF 2025

The DICE Alliance, a collaboration of ACHILLES, ACCOMPLISH, DATAPACT and CERTAIN — four Horizon Europe projects — has presented a session entitled “Semantics & Compliance: Enabling Trustworthy Data/AI in European Data Spaces”, during the European Big Data Value Forum (EBDVF 2025).

EBDVF was held in Copenhagen from 12 to 14 November 2025, gathering more than 450 experts, researchers and innovators in Data and AI.

The session was filled with over 50 participants, highlighting the growing interest in how semantic technologies and compliance-by-design can support the creation of trustworthy, interoperable, and human-centric data ecosystems.

The session opened with Dumitru Roman (SINTEF), Fenareti Lampathaki (Suite5 Data Intelligence Solutions), Rafael Rodrigues (INESC TEC), and Silvia Colella (PNO Innovation), who jointly presented how the DICE Alliance projects are translating complex European regulatory frameworks — including the AI Act, GDPR, and Data Act — into practical, machine-interpretable solutions.

The discussion then continued with Mark Dietrich (CTO, SAGE Data Spaces ), Christoph Mertens (Head of Adoption, International Data Spaces Association), and Professor George Konstantinidis (University of Southampton), who took part as end users and ecosystem representatives. They initiated a rich dialogue with project partners to explore how the DICE Alliance results align with Europe’s digital policy roadmap and how these innovations can foster adoption within real data ecosystems.

The debate, moderated by Parul Chaudhary, underscored a shared vision: that compliance and semantic interoperability are not regulatory burdens, but strategic enablers for building trust, transparency, and accountability into the next generation of AI and data systems.

During the session, ACHILLES presented its Semantic Stack, a layered architecture that bridges semantics and compliance through:

• Specification Layer defining ethical and regulatory principles in machine-readable form,
• Reasoning Layer powered by multi-agent systems for compliance assessment,
• Documentation Layer ensuring standardised, traceable metadata, and
• Governance Layer enabling auditing, accountability, and certification.

This interactive session successfully connected technical innovation with policy and industry needs — a tangible step forward in shaping European framework for trustworthy and compliant AI.