ARGENTIC Officially Kicks Off in Athens: Charting the Future of Sovereign, Energy-Aware Agentic AI

22 June 2026

The journey toward Europe’s next generation of trustworthy, decentralised artificial intelligence officially began in Athens on 18–19 June 2026. Hosted by Project Coordinator MAGGIOLI SpA at “The Golden Age” Hotel, the ARGENTIC Consortium Kick-Off Meeting brought together the project’s 20 partner organisations from 9 countries in a dynamic hybrid setting.

Over two days of intensive collaboration, researchers, technical innovators and domain experts aligned their vision to solve one of the most pressing challenges in AI today: how to build powerful, autonomous and energy-efficient AI agents operating seamlessly across the Cloud–Edge–IoT continuum.

A vision beyond centralised AI

As Generative AI rapidly transforms industries, traditional cloud-centric infrastructures are reaching their limits, facing significant hurdles in energy consumption, data privacy and latency. ARGENTIC (Agentic Reasoning for the Generative AI-Enabled Cognitive Continuum) is funded under the EU’s Horizon Europe programme (Grant Agreement No. 101298599) to build a sovereign, trustworthy alternative.

During the opening remarks, Project Coordinator Spiros Borotis (MAGGIOLI SpA) welcomed the international consortium and set the tone for the 36-month initiative. The consensus was clear: Europe must lead the way in human-centric, explainable and sustainable AI that operates where data is created, at the edge.

ARGENTIC is not just about advancing AI algorithms; it is about creating a trustworthy and energy-aware ecosystem where intelligent agents collaborate safely to serve European citizens and industry.

Key takeaways and technical foundations

The two-day meeting served as an operational launchpad, moving the project from strategic planning to concrete technical tasks. Technical leaders across Work Packages explored the core pillars of the project, focusing on the integration of lightweight Small Language Models (SLMs), Federated Knowledge Graphs and multi-objective reinforcement learning designed for energy-aware orchestration.

Partners also explored the role of Digital Twins as active validation sandboxes to simulate complex environments, facilitate privacy-preserving synthetic data sharing and target measurable reductions in carbon footprint. Rather than postponing governance to later stages, the consortium emphasised embedding explainability, fairness and compliance with the EU AI Act and GDPR into the architecture from day one.

Grounded in real-world impact: the 7 pilots

A highlight of the kick-off was setting the roadmap for user-requirement collection across ARGENTIC’s 7 large-scale pilot domains. From bedside cardiovascular monitoring in healthcare and sovereign energy management in industrial IoT, to privacy-preserving HR recruitment and immersive AR/VR cultural guides, pilot leads outlined their domain-specific expectations.

Dedicated technical workshops established the framework for the pilot-by-pilot requirement sessions planned for the summer, so that ARGENTIC’s modular architecture and SDK directly address real-world operational needs.

Looking ahead

With clear governance structures, bi-weekly technical syncs and strong synergies already forming across work packages, the ARGENTIC consortium leaves Athens with shared momentum.

As the project sets sail, the team turns its focus toward initial requirement mapping (D2.1), setting up the ethical oversight protocols and launching the project’s public outreach channels. Stay tuned as we share updates, blog posts and technical insights on our journey toward sovereign, green and trustworthy Agentic AI for Europe.