Artificial intelligence, the latest initiatives in the EU

In September 2023, during her annual State of the Union speech, president of the European Commission Ursula von der Leyen stressed the importance of technological progress for the European Union, introducing a project aimed at making supercomputers of EU agencies available to start-ups with innovative ideas. According to the European vision, the shared use of modern information technology would be a key instrument for the realization of innovative and reliable computer models: by employing artificial intelligence (AI), they would lead Europe to substantial progress in the tech development field.

AI technology, as it is employed in these models, simulates human intelligence through the creation and application of built-in algorithms in a framework of dynamic calculations. Specifically, the European Union is a pioneer in the development of supercomputers, systems aimed at developing considerable processor capabilities and executing high-performance calculation. The EU has three of the most powerful supercomputers in the world: LEONARDO, LUMI and MareNostrum 5.

Eventually, in November 2023 the European Commission and The European High Performance Computing Joint Undertaking (EuroHPC JU), responsible for the coordination of supercomputers, contributed by increasing accessibility to supercomputing resources for European start-ups dealing with artificial intelligence (AI). Access is fundamental, especially for the improvement of cloud development and scalability (the ability of increasing functions) support, since it reduces training time for using this technology from years to months (or even several weeks).

Additionally, according to President von der Leyen this initiative is a leading project for responsible AI development and its diffusion in compliance with the European values.

In December 2023, after promoting the use of supercomputers, the first EU AI Act followed: it is the first European political agreement concerning legislation aimed at regulating AI development, market placing and system usage. Basically, it makes sure that the AI systems used in the EU are in line with its rights and values, ensuring human oversight, security, privacy, transparency, non-discrimination, social welfare and sustainability. In light of potential risks related to fundamental rights, the framework indicates four levels of risk for the AI applications assessment: “minimal risk, high risk, unacceptable risk, and specific transparency risk”.   

In January 2024, after the legislation was issued, the Commission launched a package of measures aimed at supporting European SMEs (small and medium-sized enterprises) and start-ups in the development of a safe AI in compliance with EU rules. Specifically, the package facilitates the access to supercomputers for AI start-ups and SMEs through modifications of the EuroHPC Regulation, by opening AI Factories dedicated to new supercomputers. Additionally, the package includes the decision of establishing an AI office within the European Commission, coordinating European AI policies, development and supervising the compliance of the AI Act.

Further initiatives include strengthening the AI talent pool, supporting investments and accelerating the development and diffusion of “Common European Data Spaces”.

The European Commission aims at establishing communication channels and financial support measures through Horizon Europe and the Digital Europe program, involving generative AI, that will receive public and private funding until 2027 for a total of approximately €4 billion. At the same time, the Commission is establishing two European Digital Infrastructure Consortiums: the ‘Alliance for Language Technologies' (ALT-EDIC), focused on the development of common linguistic infrastructures, and the CitiVERSE EDIC, encouraging the application of AI tools to help local communities.   

Waiting for the European Parliament and Council to revise the latest package of measures introduced by the Commission, it is important to note that Europe has quickly made significant progress in the AI field and innovation, while supporting the development of this sector all around the world.

Mondo Internazionale APS – All rights reserved ® 2024