As COP30 closes, global attention is fixed on calls from more than 80 countries for a fossil-fuel phase-out roadmap. The summit has also placed artificial intelligence under scrutiny, with debates centred on the growing electricity needs of data centres and the carbon footprint of digital infrastructure. Yet one of the most direct opportunities to reduce emissions in the near term received little attention: the role of operational AI in managing energy use in commercial buildings.
Buildings remain one of the largest and slowest-moving contributors to climate change. The sector is responsible for nearly 30% of global CO₂ emissions, driven largely by heating, ventilation, and air-conditioning systems. While policymakers continue to debate the regulatory framework for AI and assess its increasing electricity use, energy specialists warn that the conversation is overlooking where AI can already deliver measurable emissions reductions.
This concern was underlined by Exergio’s CEO, Donatas Karčiauskas, who argues that AI is too often discussed exclusively as a new energy burden. In contrast, the same machine-learning methods that power data-intensive applications can also reduce demand where it matters most: inside commercial buildings.
Operational AI
Operational AI in buildings refers to software that adjusts energy consumption in real time. It analyses occupancy patterns, indoor and outdoor conditions, and the performance of HVAC equipment. It can slow or shut down output in unused spaces, shift cooling loads to off-peak periods, and respond immediately to changes in weather. Because HVAC systems represent the largest share of building-sector emissions, these optimisation measures deliver significant reductions in demand.
Sites using operational AI have recorded HVAC energy savings of 20–30%, with some large retail complexes reporting cost reductions exceeding €1 million annually. These results indicate that digital optimisation can be implemented without major construction work or equipment replacement, offering an alternative to the long timelines often associated with building-sector decarbonisation.
AI-driven optimisation
Despite this evidence, AI-driven optimisation remains largely absent from national climate strategies and from COP30’s policy discussions. Negotiators have focused heavily on supply-side transitions, including renewables, fossil-fuel phase-out, and climate finance, but demand-side measures in buildings are still underrepresented. This gap persists even though the summit’s technical sessions repeatedly highlighted the need to accelerate digital adoption and scale financial tools that support efficiency improvements.
Energy experts now argue that operational AI should be formally recognised as a decarbonisation pathway. They propose integrating building-optimisation targets into national energy-performance standards and enabling verified software-based savings to qualify for carbon-credit schemes. Such steps would align climate policy with technologies that are already reducing emissions in commercial real estate.
Digital tools
Karčiauskas notes that AI will continue to consume electricity, but its potential to reduce overall demand is significant. By lowering the energy use that forces grids to rely on fossil-fuel backup plants, operational AI can help balance the impacts of data-centre growth. The choice is not between AI expansion and climate action; rather, it is about directing digital tools to the sectors where they can deliver immediate and measurable results.
As countries prepare their post-COP30 implementation plans, the building sector’s digitalisation should be a priority. Operational AI offers a practical and scalable route to cut energy waste in buildings, bridge the gap between climate ambition and action, and support the wider transition to low-carbon systems.
Read next: Fighting cybercrime to protect digital technologies – Build in Digital
Are you a building professional? Sign up for a FREE MEMBERSHIP to upload news stories, post job vacancies, and connect with colleagues on our secure social feed.
