AESIA: A pillar in Spain’s AI regulation

Spain has taken a bold step towards the future with the creation of the Spanish Artificial Intelligence Supervision Authority (AESIA). This new body has been established to provide a regulatory framework that ensures the ethical and responsible use of AI.

A

ESIA will not only oversee compliance with national regulations but will also work closely with European Union guidelines, seeking to balance technological innovation with the protection of fundamental rights. Its mission is clear: to minimize the risks associated with AI, ensuring that its development benefits society as a whole. 

The creation of AESIA follows the approval in May 2024 of Spain’s AI Strategy by the Council of Ministers, which sets out a clear path to capitalizing on the benefits of AI while mitigating its risks. This strategy aligns with the European AI Regulation, which aims to balance technological innovation with the protection of fundamental rights and values. 

Key areas of Spain’s AI strategy 

Spain’s AI Strategy is built around three fundamental pillars: 

1. Strengthening the key drivers for AI development: This includes investing in public infrastructure for language models and fostering talent in this field. 
2. Facilitating the expansion of AI in the public and private sectors: Promoting innovation and cybersecurity, with a focus on the practical application of AI in various fields. 
3. Developing transparent, responsible, and humanistic AI: This involves promoting ethical standards and best practices in the use of AI. 

AESIA’s objectives 

The Spanish Artificial Intelligence Supervision Authority will be a key component of Spain’s AI Strategy, with five fundamental objectives: 

• Think tank and do tank. AESIA will act as a think tank and do tank, anticipating trends in the field of AI. This will include disseminating knowledge about its applications and fostering active social debate about its possibilities and limitations. 

• Supervision and evaluation. The authority will be responsible for overseeing AI systems in use, coordinating its work with market surveillance authorities. In addition, it will promote standards and best practices, ensuring that AI models comply with established regulations. 

• Regulatory coordination. AESIA will work closely with the European AI Office and other national authorities to ensure uniform application of the regulation. Its aim is to adapt regulation to technological and market changes, ensuring its effectiveness. 

• Promoting innovation. The authority will be dedicated to promoting innovation in the AI sector, facilitating compliance with regulations in both the public and private sectors. This will include the development of best practices and the implementation of sandbox environments that allow companies to experiment with new technologies without risk. 

• Global participation. AESIA will also play an active role in the global debate on AI, participating in international forums to establish minimum global standards. It will advocate for AI that is reliable, responsible, and humanistic, aligning regulation with the vision of Spain’s AI Strategy. 

Key proposals and initiatives of AESIA 

• AI certification system: AESIA will establish a certification system based on the risk levels of AI systems. This certification will oversee compliance with requirements for high-risk systems and ensure the transparency of those considered to be of minimal risk. 

• National legislation on sanctions: Although the European regulation establishes general frameworks, AESIA will work on developing a national law that defines the levels of sanctions for non-compliance and details the mechanisms of action of the authority. 

• Regulatory sandbox: The creation of a sandbox environment is being considered, which will provide companies with a controlled space to experiment with AI technologies, incentivizing innovation without the usual regulatory risks. 

Impact on Qualified Trust Service Providers (QTSPs) 

AESIA will implement a certification system similar to the existing model for Qualified Trust Service Providers (QTSPs), meaning that technology companies will be subject to strict audit processes before they can market their AI solutions. This will ensure that the use of AI aligns with ethical and security principles. 

In addition, AESIA will coordinate its actions with other authorities that affect QTSPs, such as the Bank of Spain or the Data Protection Authority, avoiding duplication of compliance processes and facilitating adaptation to new AI requirements. 

While external bodies are being accredited to issue certifications, AESIA will temporarily assume the role of authorizing the conformity of AI systems, which will streamline the process for companies already operating under the QTSP model. 

Resources and budget 

AESIA will have a team of 60 employees and an independent budget, not linked to the fines imposed. This approach reinforces its commitment to not being a punitive authority, but rather an entity that supports the responsible development and effective regulation of AI in Spain. 

The creation of AESIA represents a significant step towards the effective regulation of artificial intelligence in Spain. With its focus on transparency, ethics, and innovation, AESIA will not only protect fundamental rights and values but will also foster an environment where AI can thrive in a safe and responsible manner.