• The technical guides for understanding the high-risk requirements developed and improved during the Sandbox are already published and available for consultation.
• The Secretary of State for Digitalization and Artificial Intelligence, María González Veracruz, carried out the institutional closure with the delivery of diplomas to the 11 participating systems that have completed the sandbox and to the experts who have accompanied them throughout the process.

The Ministry for Digital Transformation and Public Service has held today the closing event of the first Regulatory Sandbox of artificial intelligence of the European Union. This event puts an end to a pioneering initiative on a continental scale, which has been aimed at generating guidelines and practical guides that facilitate companies, in particular SMEs and startups, the process of adaptation to the requirements of the European Artificial Intelligence Regulation (AI Act), the world's first comprehensive AI legislation.

The AI Sandbox, developed in collaboration with the European Commission, was born as a controlled testing environment to accompany companies.

Public-private and European cooperation

After receiving 44 applications from 18 SMEs, 6 startups, the program has guided its participants through three structured phases: training and consulting on regulatory obligations, analysis and adaptation of AI systems, and verification of compliance with the AI Act.

The Sandbox strengthens cooperation between the different actors involved at European level and is configured as an initiative open to participation and monitoring by other Member States.

Closing day

The day was attended by representatives of the AI Office of the European Commission, closed by María González Veracruz, Secretary of State for Digitalization and Artificial Intelligence and participated in the opening of the General Director of Artificial Intelligence Aleida Alcaide.

“The realization of this Sandbox allowed us to get ahead and show that it is possible to innovate while we regulate. Today that work is even more valuable because all the participants have turned this project into much more than a theoretical exercise: you have transformed it into a practical experience that shows that we can develop and deploy artificial intelligence in safe and trusted environments,” said María González Veracruz.

The Secretary of State also stressed that this initiative represents a strategic opportunity for Spain and its business fabric. “Regulation is not a brake on innovation; it is a lever that can bring competitiveness and build confidence. Deciding what technology we want is also deciding what kind of society we want to be,” he said.

In this sense, he defended a model of technological development based on responsible innovation. “The difference is not only in adopting artificial intelligence, but in how we adopt it. A competitive and innovative AI, but also safe and reliable, is precisely what we are celebrating here today,” he concluded.

Spain is ahead of European regulations

Spain has played a leading role in the construction of the European ecosystem of responsible artificial intelligence. During his presidency of the EU Council in 2023, he decisively pushed for the adoption of the European AI Regulation. She was also a pioneer in the creation of the Spanish Agency for the Supervision of Artificial Intelligence (AESIA), the first agency of its kind in Europe. The AI Sandbox, framed in the Digital Spain Agenda 2026 and the National Artificial Intelligence Strategy, is the third major milestone of this leadership.

The technical guides developed and enriched throughout this Sandbox provide a practical reference to facilitate compliance with the European AI Regulation in high-risk systems, while approving the corresponding standards. These guides are already available for consultation at: https://aesia.digital.gob.es/es/guias

This initiative to develop regulated test environments that foster innovation has been taken up in the AI bill, passed by the government on May 26, which allows the creation of additional AI sandboxes as long as they are created by market surveillance authorities or notifier, and associated with their supervisory sector. And it specifies that, in all sandboxes, the authorities responsible for defining public policies in the sectors covered by the sandbox, and the authorities of related fundamental rights must participate.

Next steps

The Agency for the Supervision of Artificial Intelligence (AESIA) presented during the closing ceremony the road map for the following phase: the dissemination of the technical guides and good practices developed during the program, and the opening of new calls for proposals. sandboxes within the framework of the AI bill passed by the Government on May 26.