Buildings account for roughly 37% of global energy-related carbon emissions, according to Schneider Electric, and new research from the company suggests artificial intelligence could help shrink that footprint. Recently published research found that adding an AI layer to a building management system can cut whole-building energy use by up to 22% compared with traditional controls, a gain Schneider estimates at about $50,000 per year at current commercial rates.
The findings are notable because conventional upgrades have fallen short. Schneider reports that smart controls, such as occupancy sensors tied to a building management system, typically reduce HVAC and lighting demand by 10% to 13%. That is helpful, but not enough to reach the 20% energy reduction by 2030 called for in many net-zero pathways. The AI configuration in the study, which was cloud-hosted, delivered reductions of 21.7% to 22.4% in a typical modeled office building, more than doubling the savings from smart controls alone.
The reason is that building conditions rarely stand still. Occupancy patterns shift, weather changes, and equipment performance drifts over time, yet control strategies often stay largely the same. Schneider says AI can connect siloed data sources, weigh forecasts and equipment performance, and adjust HVAC operation in real time. The company adds that small and mid-sized buildings under 100,000 square feet could benefit significantly, since sophisticated energy management has historically been too complex or costly for facilities with limited in-house expertise.
Those are encouraging developments, and they come with a caveat worth stating plainly. Schneider’s numbers are modeled from the first year of pilot deployments, and they describe how efficiently a building runs its heating and cooling equipment. They do not change how much heating and cooling the building needs in the first place. That second question is answered by the building envelope.
A building’s walls, roof, and foundation determine how much conditioned air escapes and how much outdoor heat, cold, and moisture gets in. Air leakage and poor insulation force HVAC equipment to work harder no matter how intelligent its controls may be. Even the best algorithm can only optimize around the load the envelope hands it. Spray foam insulation addresses that load directly by insulating and air sealing in a single application, filling gaps and cracks that other materials can leave open.
This is why the envelope and the control system should be viewed as partners rather than competitors. A tighter, better insulated building has a smaller and steadier heating and cooling load, which gives an AI-driven system a more stable and efficient platform to manage. Controls can then fine-tune the savings that the envelope has already locked in. For owners of smaller commercial buildings, the segment Schneider says could benefit significantly from AI, pairing the two approaches may offer a compelling return.
The 2030 efficiency targets are ambitious, and no single technology will close the gap alone. Building owners, contractors, and designers should treat AI-enabled controls as one promising tool alongside a high-performance envelope that cuts waste at the source.