Francesca Campolongo, the Joint Research Centre Director of Digital Transformation and Data, shares how the European Commission is working on artificial intelligence to inform policies that affect us all
As the European Commission’s science and knowledge service, the Joint Research Centre (JRC) is supporting the vision of Europe in becoming a leader in artificial intelligence (AI) and adopting innovative AI legislation. Here, I share how the JRC is working on AI to inform policies that affect us all.
Developing European Union (EU) digital and AI policies poses significant challenges, as policy must navigate complexity, innovation needs, ethical dilemmas, and global competition. Science plays a crucial role in supporting the whole policy process with sound scientific evidence. It anticipates and mitigates risks, such as those related to safety and bias, providing a solid grounding as the technological landscape evolves.
With the recent launch of the Apply AI Strategy and AI in Science Strategy, the EU made another step towards making the most of the opportunities offered by AI and boosting its adoption in strategic sectors. JRC research contributed to the early stages of both these initiatives. JRC science provided insights in key areas identified by the Apply AI Strategy, on AI uptake by businesses and the public sector, and on its impact on the educational offer and job demand.
This evidence is very important. The job market is evolving rapidly, with AI reshaping the skills in demand. Similarly, businesses – and in particular SMEs – and public administrations all over Europe need to implement and boost AI solutions. Additionally, JRC provided the scientific and technical foundation for the European Strategy for AI in Science, offering a detailed analysis of the use of AI in the scientific process and the landscape of AI in science.
This evidence serves policies to unlock AI’s full potential for EU research and industrial sectors, ultimately impacting people’s lives. Thanks to JRC, for instance, public servants can now find practical, actionable solutions to leverage AI use in their administrations. JRC assessment of the European Digital Innovation Hubs will be functional to improve their services in support of business digital innovation.
This will allow entrepreneurs to access better services and make informed AI-investment decisions. JRC mapping of AI’s education offer against labour demand will benefit policymakers and educators dealing with AI-training offers and closing skill gaps. Lastly, our analysis of AI in science will help scientists learn about the gains and risks of embedding AI in their scientific work.
Supporting the most difficult steps of AI lawmaking
JRC has supported the design of AI regulation from the start, including the groundbreaking AI Act. The new Act, the first of its kind, needed to balance innovation with safety and ethics, and be future-proof. Not easy in such a fast-changing field! Once again, the role of science was key in dealing with the complexity of this transformative technology, providing evidence where it was needed the most. In this case, JRC experts drafted technical terminology definitions, such as AI system versus AI model, generative AI, and general-purpose AI, and provided ad-hoc analyses on areas such as high-risk AI systems, general-purpose AI, and transparency criteria. These definitions were crucial to avoid legal loopholes, ensure technical soundness, and help to balance regulatory requirements – not too broad, not too narrow.
Opening the black box to defend users’ online rights
Research remains still an essential component when we need to put policy into action and assess whether the activities have reached their intended goals. An excellent example of the value of science at this stage of policymaking is the work of our European Centre for Algorithmic Transparency (ECAT).
ECAT scientists are supporting the implementation and the enforcement of the Digital Services Act, “opening the black box” of AI algorithms. By producing cutting-edge scientific and technical evidence on the societal risks of algorithmic systems and AI, scientists empower policymakers to address complex online harms. This includes investigating how platform design, recommender systems and AI tools may affect the mental and physical health of minors.
Their work is already driving measurable positive change across major platforms used by people across the EU, such as TikTok and Meta, demonstrating exceptional innovation, impact, and policy relevance at every step. Some online platforms introduced changes to their systems and interfaces, such as X’s feature to report illegal content or increased transparency with options to disable the personalised feed.
Implementing AI policy with the JRC Scientific AI Hub
Looking ahead, the implementation of AI in the Science Strategy will also rely on the upcoming JRC Scientific AI Hub. This hub will serve the need to assess the performance of AI models and systems for strategic scientific research and, by design, will work hand in hand with policymakers at the European AI Office.
Embedding science into policy design is strategic, and science is increasingly recognised as an essential part of the policy process. The Joint Research Centre will stand at the forefront of science-driven AI policy. We are ready to build on this momentum and draw on our expertise to ensure that the sciences will remain the compass for sound policymaking.
