Blog

Shaping the Future of AI Education and Industry: Insights from the MERIT International Workshop 

On June 18, 2025, the MERIT project hosted the Hybrid International Industry Engagement Workshop on Artificial Intelligence, drawing experts, educators, entrepreneurs, and policymakers together from across the globe. The event, organized by the MERIT consortium, aimed to bridge academic innovation and industry needs by refining AI-focused master’s programmes through active engagement with the private sector. This workshop revealed not only the urgency of integrating AI into higher education curricula but also the strategic foresight required to do so responsibly and effectively. Below are the key themes and insights extracted from the workshop. 

AI in Healthcare: A Data-Centric Revolution 

Giuseppe Jurman, from FBK Trento and the Humanitas University of Milan, delivered the opening keynote, setting a compelling tone. His talk underscored the growing importance of data science in healthcare and emphasized the critical role AI now plays in medical research and clinical practice. Highlighting deep learning and synthetic data generation, Giuseppe Jurman presented a future where clinical trials may rely on AI-generated virtual patients. He warned, however, that effective AI in healthcare hinges not merely on algorithmic power but on rigorous data management —“data cleaning and wrangling consume up to 70% of our project time” he noted. Importantly, he addressed ethical challenges and regulatory constraints such as GDPR, calling for a balance between data utility and patient privacy. 

Four Ways AI Projects Fail — and How to Succeed 

In a compelling and practical session, Egidijus Pilypas, co-founder of Exacaster, unpacked the reasons why 74% of AI initiatives fail to deliver business value. His presentation offered a refreshingly candid look at AI adoption pitfalls: 

1. Lack of strategic intent – implementing AI out of fear or trendiness often leads to costly failures. 

2. Poor validation processes – organizations frequently invest heavily in an idea before testing it. 

3. Unrealistic automation goals – overestimating AI’s reliability without human oversight can damage customer trust. 

4. Incomplete teams – success in AI requires cross-functional teams including product managers, data engineers, and developers—not just data scientists. 

His takeaway? Start small, validate quickly, and structure AI teams like product teams for real impact. 

Immersive AI in Higher Education 

Carlos Dominguez of Torrens University in Australia showcased the cutting-edge work his team is doing to build agentic, AI-driven learning environments. These immersive systems combine Unity3D, avatars, and generative AI to personalize learning pathways and make education more interactive and inclusive. One standout example was “Safe Passage”, a gamified platform teaching cybersecurity through simulations. Another was an AI-driven HR manager avatar trained on course material to help psychology students role-play workplace scenarios. Carlos Dominguez emphasized the importance of user experience, accessibility, and backend architecture—including hybrid retrieval-augmented generation (RAG) models and secure Learning Tools Interoperability (LTI) integration. 

AI Regulation as a Catalyst for Skills Development 

Tomasz Kramer from Luxembourg introduced a novel framework called “Reg-to-Skills”, which translates EU legislation—like the AI Act—into competency-based learning paths. Rather than treating regulations as burdens, Tomasz Kramer argued they should guide curriculum design and employee upskilling. His methodology uses AI to parse legal texts, map them to business roles, and align them with frameworks like ESCO (European Skills, Competences, Qualifications, and Occupations). He urged educators and businesses to view compliance not as an obstacle, but as a strategic opportunity to build future-ready teams. 

Panel Discussion: Redefining the AI Specialist 

The workshop concluded with a panel that tackled a fundamental question: what AI skills do future professionals truly need? The consensus was that education must go beyond superficial tool usage to cultivate deep conceptual understanding. 

Panelists highlighted that AI education should: 

• Teach how to learn and adapt, not just specific tools 

• Balance black-box AI usage with knowledge of underlying systems 

• Include ethical, legal, and business perspectives 

• Integrate practical, project-based learning with foundational theory 

The discussion acknowledged AI’s broader cultural and societal implications, likening it to the invention of the printing press in its potential to reshape knowledge creation and dissemination. 

Conclusion: Toward a Human-Centric AI Future 

The MERIT International Workshop did more than identify technical trends or educational gaps—identified a shared commitment to developing AI that is trustworthy, inclusive, and aligned with human values. From healthcare to immersive education, from compliance to capability building, the workshop laid down a challenge: we must not only prepare students and professionals for AI—we must prepare AI for society. 

As Europe and the world accelerate toward digital transformation, the MERIT consortium continues to exemplify what it means to co-create the future of AI through education, ethics, and collaboration.