The rapid adoption of artificial intelligence (AI) is fundamentally changing DC design and energy requirements.  Projections indicate that AI-driven growth will substantially increase global DC electricity consumption in the EU by 2030, potentially doubling current demand or exceeding it in the absence of efficiency improvements. The predicted DCs energy use raises concerns regarding grid capacity, renewable energy availability, water use, and investment requirements. Combining large centralized DCs for AI training with smaller distributed edge DCs for inference offers a promising pathway to balance efficiency, latency, and grid integration. 

The presentation concludes that AI-enabled digital infrastructures offer transformative potential across sectors but will significantly increase energy demand. Maximizing efficiency in hardware, cooling, and algorithms, combined with access to renewable energy, enabling waste heat reuse, and supporting grid flexibility are essential to limit AI future GHG emissions. Coordinated action across EU, national, and local governance levels will be critical to ensure that future AI development delivers societal benefits while remaining environmentally sustainable