KPMG: AI cuts building energy waste by up to 30%

KPMG's new report says that using AI can cut wasted energy in buildings by up to 30%. Smart software helps control things like heating, cooling, and lighting, saving more than expensive equipment upgrades. Real-world examples show big savings in office towers, schools, and even Google's data centers. Many building managers plan to invest more in AI, but teams still need to learn how to use it well. By using AI, buildings can waste less energy and help the planet, all while saving money.

A landmark 2025 report from KPMG reveals that artificial intelligence (AI) cuts building energy waste by up to 30%. The study highlights how smart software algorithms can continuously optimize HVAC, lighting, and power systems. These findings suggest that software-driven efficiency offers deeper savings than costly hardware retrofits, prompting a shift in capital planning for facility leaders.

Key Findings from the KPMG AI Energy Report

Artificial intelligence systems reduce building energy waste by analyzing real-time data from sensors and autonomously adjusting heating, cooling, and lighting controls. This continuous optimization fine-tunes equipment performance every minute, eliminating inefficiencies that manual oversight or scheduled retrofits would otherwise miss, leading to significant energy savings.

KPMG's analysis of numerous commercial buildings demonstrated an average 20-30% reduction in energy waste when AI was integrated with strategic energy management platforms. The firm emphasizes that these savings stem from operational improvements, not immediate hardware replacement. Given that buildings contribute to nearly 40% of global greenhouse gas emissions, according to a LightNOW summary, even a partial reduction helps organizations meet net-zero goals. KPMG advises a phased approach to maximize ROI:

Tier 1: Optimize existing systems with continuous fault detection and automated control adjustments.
Tier 2: Replace underperforming or old assets identified by AI analytics.
Tier 3: Implement on-site renewables after minimizing the building's energy load.

The report cautions that adopting renewables before achieving operational efficiency leaves "cheap carbon on the table."

Real-World Examples of AI-Driven Energy Savings

The report's findings are validated by numerous real-world applications. For instance, a 32-story office tower in Manhattan deployed BrainBox AI and cut its HVAC energy consumption by 15.8% in less than a year, saving over $42,000 and 37 metric tons of CO2e, as reported by Time Magazine. This was achieved through a software-only implementation, avoiding any disruption to tenants.

Other sectors show similar success:

Education: The Stockholm school district reduced electricity use by 8% across 87 facilities by adding an AI management layer.
Technology: Google's DeepMind project cut data center cooling energy by as much as 40%.

This trend is confirmed by market data; a Honeywell survey revealed that 84% of facility managers intend to increase their AI budgets, with 55% already using AI for energy management link.

Overcoming Implementation Challenges

Adopting AI in building management is not without its hurdles. Key challenges include:

Skills Gap: The 2025 Workplace Index found that 50% of organizations do not have the in-house expertise to act on AI-driven insights.
Data Fragmentation: Siloed systems for HVAC, lighting, and access control prevent the creation of a unified data model essential for effective AI.
Governance and Security: Predictive algorithms require access to sensitive occupancy and operational data, raising privacy and security concerns.

KPMG recommends a strategic approach: begin with a pilot project, create cross-functional teams of facility engineers and data scientists, and establish a monthly feedback loop to refine the algorithms. Treating AI as an evolving system, rather than a one-time installation, is crucial for long-term success and scalability.

Ultimately, software-led optimization is the key to unlocking nearly a third of the energy currently wasted in commercial buildings. This approach not only reduces operating costs and carbon emissions but also establishes a new standard for efficiency. Organizations that successfully integrate their people, processes, and data with AI are already realizing these benefits and leading the way into 2026.