Background
With the entry into force of the EU AI Act (Regulation 2024/1689), AI systems used in critical infrastructure, including energy supply networks are subject to stringent requirements regarding data governance, logging, and human oversight. Operators of AI-powered grid, market, and flexibility services must be able to demonstrate compliance with these requirements. Simultaneously, technical platforms are emerging to facilitate the exchange of data for these AI services. While these systems already generate extensive metadata such as policies, contractual commitments, and usage logs, etc. there is currently no link connecting these logs to the specific articles of the AI Act. Without such a link, compliance remains a manual, document-driven process that fails to scale with the dynamic nature of data-driven AI services. This work develops the conceptual and prototypical foundation for automated compliance tracking and is intended as an independent publication.
The overarching goal is the prototypical implementation and evaluation of a compliance-tracing framework that automatically analyzes connector logs—within an EDC- or FIWARE-based energy data space against the requirements set out in the articles of the AI Act, ultimately generating an auditable compliance report.