ENFIELD presents the Final Conference 2026 – “Advancing AI in Europe: From Foundations to Real-World Applications”

Two-day hybrid event in Brussels connected leading voices from research, industry and public policy to examine how Europe can translate AI excellence into sustainable, trustworthy and impactful applications.

More than 200 participants joined the ENFIELD Final Conference, “Advancing AI in Europe: From Foundations to Real-World Applications,” held on 7–8 July 2026 at the Palais des Académies in Brussels and online.

The two-day hybrid event brought together researchers, industry leaders, policymakers, European institutions and representatives of major AI initiatives to discuss how Europe can move advanced artificial intelligence research beyond the laboratory and into practical applications that create economic and societal value.

Across keynote presentations, expert roundtables, demonstrations and networking sessions, the conference addressed the technical, environmental, regulatory and strategic conditions required for the responsible deployment of AI at scale.

The programme reflected a central challenge for Europe: turning strong research capabilities into AI systems that are robust, sustainable, secure, trustworthy and suitable for real-world use.

Day One: Sustainable Infrastructure, Robustness and Trustworthy AI

The first day opened with a keynote by Rod Evans of NVIDIA on Green AI and Sustainable Infrastructure, with a particular focus on green data centres.

The presentation examined the resource and infrastructure demands associated with large-scale AI. As increasingly complex systems require greater computing capacity, the environmental impact of data centres and digital infrastructure has become a critical concern for Europe’s technological development.

A roundtable on Edge AI: Robustness, Continual Learning and Real-World Deployment then explored the challenges of deploying AI closer to devices, users and operational environments.

Contributors included Antoine-Alexandre André of the European Association for Artificial Intelligence, Eli Tylman of KU Leuven, Ovidiu Vermesan of SINTEF and the European Alliance for Edge AI, and Andon Tchekmedjiev of IMT.

The discussion addressed how edge AI systems can remain effective when operating with limited computing power, memory, energy and connectivity. Speakers also considered continual learning, system adaptation and the need to maintain reliability as operational conditions change.

The following roundtable, Sustainable AI — Challenges and Directions for Future AI Applications, brought together Jonas Mühlberg of Université libre de Bruxelles, Martin Willame of UCLouvain, Alexandra Papageorgiou of KU Leuven, Nicki Lisa Cole of KNOW, and Erik Johannes Husom of SINTEF.

The panel expanded the sustainability discussion beyond infrastructure by examining the environmental, social and organisational impact of AI throughout its lifecycle. Participants considered how sustainability can be incorporated into research priorities, procurement, system evaluation and deployment decisions.

The afternoon programme focused on how broad principles of responsible AI can be converted into concrete industrial practices.

The session From Principles to Practice: Assessing Trustworthy and Sustainable AI in Industry included contributions from Paolo Giudici of OECD.AI, Nicolas Rebierre, Alexander Machado and François Picard of DTI.

The panel examined how organisations can assess AI systems against criteria such as transparency, accountability, fairness, safety and sustainability. Speakers discussed the role of evaluation frameworks, measurable indicators, documentation, risk assessments and governance procedures.

The roundtable AI Trustworthiness in the Wild: Closing the Reality Gap continued this discussion by focusing on the difference between AI performance in controlled conditions and its behaviour in real-world environments.

The session featured Sebastian Heil of TUC, Giovanni Bonati of MAG, Georgios Spathoulas of NTNU, and Javier Augusto of BAS.

Speakers explored how changes in data, context, user behaviour and operating conditions can affect system performance. They also addressed the importance of security, explainability, resilience, validation and human oversight.

The first day concluded with the roundtable AI Research Funding in Europe: From Networks of Excellence to the AI Continent Action Plan and Beyond.

Participants included Alain Pagani of DFKI, Cecile Huet of the European Commission, Georgios Spathoulas of NTNU, and Ana Solange Leal of INOVA+.

The session examined how European funding programmes and Networks of Excellence can strengthen research collaboration, connect institutions across borders and accelerate the transfer of scientific advances into practical applications.

The debate also placed ENFIELD’s work within Europe’s wider ambitions to develop a competitive, responsible and strategically autonomous AI ecosystem.

Day #1 – 07 June 2026

Day Two: Skills, Strategic Capacity and Europe’s AI Future

The second day opened with a keynote by Virginia Dignum of Umeå University on The AI Paradox: A Guide for Responsible Humans.

The presentation focused on the relationship between increasingly capable AI systems and the responsibilities of the people and institutions that design, deploy and govern them. It placed human agency at the centre of responsible AI, emphasising that technological progress does not remove the need for judgement, accountability and democratic oversight.

A roundtable on Explainability and Human Oversight in AI Systems then explored how organisations can ensure that AI-supported decisions remain understandable and subject to meaningful human control.

Contributors included Johan Loeckx of VUB, Isel Grau of TU/e, Jeriek van de Abeele of Telenor, Géza Németh of BME, and Mohammed Al-Radhi of BME.

The discussion considered the role of explainability in helping users, operators and affected individuals understand how AI systems reach their conclusions. Speakers also examined the limits of technical explanations and the need to adapt information to different audiences, contexts and levels of expertise.

Human oversight was presented not simply as the presence of a person in the decision-making process, but as the ability to question, challenge and override an AI system when necessary. The session underlined the importance of clear responsibilities, appropriate training and interfaces that support effective intervention.

The following roundtable, Code, Compliance, and Competitiveness: Making the AI Act Work for Europe, brought together Giovanni Bonati of Maggioli, Jerry Spanakis of Maastricht University, Julia Palma of CeADAR, Astik Samal of Maggioli, and Alberto Casanova of Maggioli.

The panel examined how the requirements of the EU AI Act can be translated into practical processes for organisations developing and deploying AI. Participants considered issues such as risk classification, technical documentation, transparency, conformity assessment and ongoing monitoring.

A central theme was that regulation and competitiveness should not be treated as opposing objectives. Effective implementation of the AI Act could strengthen trust, reduce uncertainty and create more consistent conditions for the development of responsible AI across Europe.

The afternoon programme shifted towards the people, skills and strategic capabilities required to support Europe’s long-term AI ambitions.

The session Shaping Europe’s AI Talent: Literacy, Training and Future Skills included contributions from David Timis of the Global Governance Institute, Thérésa Elhage of All Digital, Adina Florea and Ștefan Trăușan-Matu of UPB, and Kristina Vassiljeva of TalTech.

The panel considered how Europe can develop the technical, managerial and societal skills needed for the wider adoption of AI. Participants discussed the importance of AI literacy not only for researchers and engineers, but also for employees, public officials, educators and citizens.

The session reinforced that Europe’s AI capabilities will depend not only on infrastructure and research investment, but also on whether people are equipped to use, evaluate and govern AI systems responsibly.

The roundtable EU AI Sovereignty: A Two-Fold Challenge continued the debate by examining Europe’s ability to develop and control critical AI capabilities.

The session featured Valerie Zapico of Valkuren, David Timis of the Global Governance Institute, Ricardo Bessa of INESC TEC, Pankaj Pandey of NTNU, and Jeriek van den Abeele of Telenor.

Speakers explored the technological and political dimensions of AI sovereignty. The discussion considered Europe’s dependence on external providers for computing infrastructure, advanced models, data platforms and other strategic resources.

The panel also considered the risks of interpreting sovereignty as isolation. Stronger European capabilities would still require international cooperation, open research networks and partnerships with industry, while reducing critical dependencies and strengthening resilience.

The second day concluded with the roundtable From Lighthouse to Continent: ENFIELD and the Future of European AI Leadership.

Participants included Habtamu Abie of Norwegian Computing Center, Adina Florea of UPB, Ricardo Bessa of INESC TEC, Adam Brandstetter-Kunc of ESF, and Leïla Belmerhnia of Predict.

The session reflected on how the knowledge, networks and practical results developed through ENFIELD can contribute to Europe’s wider AI ecosystem. Speakers considered how successful research initiatives can move beyond individual projects and create lasting cooperation between scientific institutions, industry, public authorities and innovation communities.

The closing debate positioned ENFIELD as part of Europe’s broader effort to move from isolated centres of excellence towards a connected and sustainable continental AI ecosystem. It reinforced the need for continued investment, shared infrastructure, coordinated policy and long-term collaboration if Europe is to achieve responsible and competitive leadership in artificial intelligence.

Day #2 – 08 June 2026

ENFIELD’s Strategic Outcomes and Roadmap

As its final conference, the event provided an important platform for ENFIELD to present its strategic outcomes, lessons and roadmap for the future.

ENFIELD highlighted the role of European Networks of Excellence in building durable research communities, supporting interdisciplinary collaboration and creating stronger pathways from scientific advances to practical impact.

The conference underlined that Europe already possesses strong research capabilities and a growing policy framework for responsible AI. The next challenge is to connect these strengths more effectively with investment, infrastructure, skills and real-world adoption.

Demonstrations and Cross-Sector Exchange

Demonstration sessions held throughout the conference enabled participants to explore technologies, tools and research results in a more practical setting.

The in-person participants explored demo videos and now more than 50 videos showcasing ENFIELD-supported research, technologies and real-world applications. The complete collection is also available on the ENFIELD YouTube channel.

YouTube Playlist: ENFIELD DEMO VIDEOS

Conference Recordings

Recordings from the ENFIELD Final Conference will allow participants and the wider AI community to revisit the keynote presentations, roundtables and strategic discussions from both days.

Day #1 recording: ENFIELD Final Conference – Day #1 – 07 July 20