UBITECH announces its participation in AICONFORM — “Agentic Framework for Automated Regulatory Compliance” — a new action funded under the Digital Europe Programme (Grant Agreement No. 101298515), which was officially launched at a kick-off meeting hosted in Athens by the Ministry of Digital Governance of the Hellenic Republic, the project’s coordinating partner. The project runs for 24 months, having started in June 2026. Responding to the DIGITAL-2025-AI-08-COMPLIANCE call, AICONFORM sets out to deliver a modular, interoperable framework that automates the extraction, exchange, verification, and auditing of regulatory obligations across sectors. The framework operationalises core pieces of the EU’s digital rulebook — the AI Act, NIS2, and eIDAS 2.0 — through machine-readable templates, a legal knowledge graph, and agentic workflows capable of translating legislative updates into executable compliance checks. Privacy-preserving technologies, including homomorphic encryption and differential privacy, are embedded alongside qualified trust services such as qualified electronic signatures, e-seals, and timestamping, with integration of the EUDI Wallet to ensure secure identity management and non-repudiation.
The framework will be validated across three real operational environments spanning public administration, manufacturing, and energy, demonstrating end-to-end compliance reporting, incident monitoring, and conformity documentation under realistic conditions. By the end of the project, the consortium targets a reduction of at least 30% in reporting time, reporting accuracy of at least 95%, a minimum of five machine-readable templates per pilot, and an open-source toolkit with documented APIs and SDKs aligned with SIMPL and IDSA specifications, laying the groundwork for replication and uptake by SMEs and public administrations across Europe.
Within the AICONFORM consortium, UBITECH contributes a broad and technically demanding portfolio of work. The company will translate legal knowledge graph outputs into executable rules in a Drools-style engine, complemented by a human-friendly editor for parameterisation and versioning, and will implement explainability hooks providing rationale and traceback for every automated determination. UBITECH will also integrate Article 6 risk-classification logic under the AI Act, benchmark rule-evaluation throughput against pilot-representative workloads, and define acceptance criteria for latency and correctness. Explainability will be reinforced through LIME and SHAP-based techniques, ensuring that each compliance determination is accompanied by a traceable rationale.
UBITECH will further implement guarded agents that combine SPARQL query chains with LLM prompts to support conformity assessment and the automated drafting of technical documentation. Working closely with the Ministry of Digital Governance, FACTOR SL, and MIWenergia, UBITECH will co-create pilot charters, documenting processes, data sources, and reporting obligations, and mapping them to AICONFORM’s capabilities and to the relevant regulatory clauses of the AI Act (including Annex III), NIS2 timelines, and eIDAS 2.0 qualified electronic signature requirements. This work will produce pilot-specific Requirement Specifications and Data Sharing Agreements, and will extend to requirements for general-purpose AI systems under Chapter IV of the AI Act where relevant.
On the privacy and governance front, UBITECH will integrate homomorphic encryption operators, such as SEAL, together with differential-privacy mechanisms to enable aggregate reporting without exposing raw underlying data, and will define governance policies enforced through platform services and metadata embedded in the knowledge graph.
UBITECH’s technical contributions to AICONFORM are led by its Big Data Engineering, Analytics & Science (DEA) Research Group (ubitech.eu/dea), while overall coordination of UBITECH’s involvement in the project is managed by the company’s Project Management Office (ubitech.eu/pmo).
“AICONFORM comes at a pivotal moment for European digital regulation, when the AI Act, NIS2, and eIDAS 2.0 are converging to reshape how organisations demonstrate compliance. UBITECH is proud to bring its expertise in knowledge graphs, agentic AI, and privacy-preserving technologies to a consortium coordinated by the Ministry of Digital Governance of the Hellenic Republic. Our objective is to help turn regulatory text into executable, auditable, and explainable logic that public administrations, manufacturers, and energy operators can trust and adopt at scale,” said Christina Stratigaki, Head of UBITECH’s Project Management Office.
