Advances in artificial intelligence (AI) in recent years have transformed industries and societies worldwide. Most of these developments have focused on the General purpose AI (IAPG), which focuses mainly on generative AI through chatbots. For this reason, the EU has invested in the Enterprise's supercomputer network European High Performance Computing Community (EuroHPC JU), which are being adapted to train large-scale IAPG models and new AI applications. These supercomputers are being optimized in ecosystems known as AI Factories.
In December 2024, the EuroHPC JU designated seven consortia to establish the prioritieseres AI Factories in Europe. These were implemented in Finland, Germany, Greece, Italy, Luxembourg, Spain and Sweden. Since then, the total has increased to 13 factories, focused on providing free and personalized support to SMEs and startups. In addition to providing services to European industry, AI Factories work to support European scientific users in developing large AI models and improving skills and knowledge in the field of AI.
AI Factories are designed to facilitate access to larger datasets, including open data. This data is essential for training AI models and chatbots, as well as for centralizing data resources and making them accessible. AI Factories are designed to act as intermediaries that connect the end user with the open datasets that are relevant to them, ultimately promoting the reuse of public sector data. These factories also enable the standardization of data formats, making open data more usable and interoperable across different AI systems and tools.
As Europe works to integrate AI ethically and innovatively into business practices, the role of AI factories is becoming increasingly important. By connecting advanced computing with open data, AI factories enable startups, researchers and public institutions to unlock the potential of AI. You can visit data.europa.eu to better understand the impact of AI factories on open data exchange, from weather analysis to benchmarking.
