Support guides for compliance with the AI Act published

A Coruña, 16 December 2025 – The Minister for Digital Transformation and the Civil Service, Óscar López, has announced the publication of guidelines to support compliance with the Artificial Intelligence Act for high-risk systems. The publications, available here, are the result of the Spanish AI Sandbox, a pioneering initiative based on the creation of a controlled testing environment to drive innovation and make it easier for SMEs, start-ups and large companies to comply with European regulations.

This is a set of 16 practical guides that aim to support the implementation and compliance of the  Artificial Intelligence Act (AI Act) and provide recommendations aligned with regulatory requirements pending the approval of harmonised implementation rules for all Member States. These manuals will help the Spanish productive sector to develop innovative, reliable AI systems that respect fundamental rights.

The guidelines, which are not binding and do not replace or develop the applicable regulations, have been structured into three sections. On the one hand, there are two guides with informative material to facilitate understanding of the RIA; on the other, there are thirteen guides on technical requirements in areas such as risk management, data governance, transparency and cybersecurity, among others; and finally, there is a checklist tool for each obligation to help suppliers and those responsible for deploying high-risk AI systems to comply with the regulations.

These documents are subject to ongoing evaluation and review, with periodic updates in line with developments in standards and various guidelines published by the European Commission, and will be updated once the Digital Omnibus amending the AI Act is approved.

AI Sandbox:

The guidelines now being published are the result of the AI Sandbox that Spain launched last April, a pioneering initiative in Europe to advance the understanding and implementation of RIA. It has been promoted by the Secretary of State for Digitalisation and Artificial Intelligence, through the Directorate-General for Artificial Intelligence, and has been supported by national market surveillance authorities such as AESIA.

In this sandbox, or controlled environment in which to test a system under real conditions before its market launch, high-level training and public consulting services have been provided to twelve high-risk AI systems from Spanish companies selected to be part of the project and operating in six different sectors: essential services, biometrics, employment, critical infrastructure, machinery and healthcare products.

The participating entities have been able to test and validate the suitability of their systems to the requirements of the Artificial Intelligence Act through conformity assessment simulations. Governance and documentation requirements such as risks, data traceability and human oversight have been tested, as well as others related to technical performance: robustness, sensitivity, demographic bias, error rates, etc.

The main objective is to support the Spanish ecosystem in meeting AI requirements, promoting competitiveness and reducing the regulatory burden, especially for SMEs and start-ups.

The 16 guidelines now being published are the result of the work, learning and experience of this regulatory sandbox and will serve to support other companies that develop, market or deploy AI systems, guiding them in compliance, implementation and actual application of the regulations. They are yet another example of the commitment of SEDIA in general and AESIA in particular to the development and use of responsible and reliable artificial intelligence that respects fundamental rights and leaves no one behind.