Consumers believe artificial intelligence will consume nearly a fifth of the world's electricity within three years, but new research suggests the reality is dramatically different. According to Bain & Company's fourth edition of the Visionary CEO's Guide to Sustainability, executives expect AI to account for around 11% of global energy consumption within three years, while consumers estimate 19%. However, Bain's own modeling forecasts a much smaller share of just 0.7%.

This perception gap matters far more than a simple misunderstanding. The disconnect between what people believe AI consumes and what it actually consumes is already influencing consumer behavior and shaping how companies approach sustainable technology. As AI infrastructure expands globally, getting this narrative right has become critical for building public trust and directing investment toward genuine sustainability challenges.

Why Are Consumers So Wrong About AI's Energy Footprint?

The gap between perception and reality is striking. Consumers overestimate AI's share of global energy consumption by nearly 30 times compared to Bain's modeling forecasts. This misalignment reflects broader confusion about how AI systems actually work and where energy consumption concentrates in the technology sector.

Part of the problem stems from how AI has been portrayed in media and corporate communications. Dramatic headlines about data centers, massive computing clusters, and the computational power required to train large language models have created an outsized impression of AI's energy demands. Meanwhile, the actual efficiency gains AI enables across other industries receive far less attention.

"AI's sustainability impact is increasingly becoming part of the consumer conversation, but there is a significant gap between perception and reality. Consumers overestimate AI's share of global energy consumption by nearly 30 times, and those concerns are already influencing how they use AI," said Wissam Yassine, Partner and Middle East Sustainability practice leader at Bain & Company.

Wissam Yassine, Partner and Middle East Sustainability practice leader, Bain & Company

The concern is real enough that it's changing purchasing decisions and technology adoption patterns. Companies now face pressure to demonstrate that their AI systems are environmentally responsible, even when the actual energy impact is minimal compared to other sectors.

How Are Leading Companies Actually Using AI for Sustainability?

Rather than focusing solely on reducing AI's own energy footprint, companies at the forefront of sustainable technology are leveraging AI to create tangible environmental and business value elsewhere. The distinction matters because it reframes the conversation from "How do we make AI greener?" to "How do we use AI to make everything else greener?"

Bain identified significant differences in how companies approach sustainable AI. Among leading "shapers," 90% see AI as a major opportunity to advance sustainability goals, compared with just 41% of lagging companies. This gap reflects a fundamental strategic choice about whether AI is viewed as a sustainability problem or a sustainability solution.

• Operational Efficiency: Leading companies use AI to improve operational and energy efficiency across manufacturing, logistics, and building management, reducing overall energy consumption far more than AI itself consumes.
• Commercial Advantage: Companies are turning sustainability into a competitive advantage by using AI to identify and market sustainable products and services to increasingly conscious consumers.
• Climate Risk Management: AI helps identify and manage climate risks by analyzing physical exposure, supply chain vulnerabilities, and transition risks that traditional financial analysis might miss.

The companies succeeding in this space are bringing together business, technology, and sustainability teams around a shared definition of value. This alignment allows them to scale use cases that deliver both financial returns and measurable environmental outcomes.

Steps to Align AI Strategy with Genuine Sustainability Goals

• Move Beyond Energy Metrics: Stop measuring AI sustainability success solely by kilowatt-hours consumed. Instead, track the environmental impact AI enables across the entire organization, including emissions reductions in operations, supply chains, and product design.
• Bring Teams Together: Create shared governance between business leaders, technology teams, and sustainability professionals. Misalignment between these groups often leads to AI investments that look good on paper but deliver limited real-world environmental benefit.
• Communicate Honestly: Be transparent about both AI's actual energy footprint and the sustainability value it creates. This builds consumer trust and differentiates companies from those relying on greenwashing claims.
• Identify Where Technology, Policy, and Behavior Align: Focus AI investments on areas where regulatory support, consumer demand, and technological capability converge, making scaling more feasible and impact more durable.

What Does the Broader Sustainability Investment Landscape Look Like?

The AI energy perception gap exists within a larger context of uneven progress across sustainable technologies. Private companies and governments invested $17 trillion in sustainable technologies over the past decade, reaching a record $2.4 trillion in 2025. However, this investment is heavily concentrated in areas where technology, policy, and consumer behavior have already aligned.

Solar, batteries, and electric vehicles are the only three of 37 sustainable technologies tracked by Bain that have outperformed forecasts made a decade ago, while 29 fell short. This pattern reveals an important lesson for AI: perception and reality matter less than whether the underlying conditions for scaling actually exist.

Agriculture, manufacturing, materials, and natural capital received less than 10% of sustainable technology investment despite accounting for around 37% of global greenhouse gas emissions. This mismatch suggests that capital is flowing toward visible, consumer-facing solutions rather than the harder, less glamorous work of decarbonizing heavy industry and food systems.

Consumer willingness to support sustainability is growing. Bain found that 85% of 7,500 consumers surveyed across the US, UK, Italy, Brazil, and Indonesia are concerned about environmental sustainability, up from 79% the previous year. Moreover, 83% have adopted at least three sustainable lifestyle habits, and consumers are willing to pay an average 18% premium for sustainable products, rising to 24% when products offer health benefits.

What Role Does AI Play in Green Finance and Sustainable Investing?

Beyond operational sustainability, AI is reshaping how investors identify and evaluate genuinely sustainable companies. AI-powered green finance uses machine learning and predictive analytics to process environmental, social, and governance (ESG) data at scale, helping analysts assess environmental and financial performance together.

The practical value is significant. A fund analyst screening companies for a "green" portfolio traditionally faced a daunting task: determining whether a company's sustainability disclosure was meaningful or merely well-written intent with no measurable substance. Multiply that question across hundreds of holdings and thousands of pages of disclosure in inconsistent formats, and manual verification becomes nearly impossible.

AI tools can now scan annual reports, sustainability disclosures, and news coverage simultaneously, flagging where the language of a disclosure doesn't match the numbers reported elsewhere. Natural language processing models can identify inconsistencies between stated commitments and actual workforce data, for example. Predictive models can combine historical climate data with a company's asset locations to estimate physical climate risk exposure that wouldn't be obvious from a balance sheet alone.

However, these tools have real limitations. ESG and climate datasets are often incomplete, inconsistently reported across jurisdictions, or self-disclosed without independent verification. Different ESG data providers can score the same company quite differently since there is no single, universally agreed scoring standard. AI models built on unreliable inputs produce unreliable outputs, regardless of how sophisticated the model itself is.

The key insight is that AI doesn't make investment decisions by itself. What it does is compress the research time an analyst needs to reach an informed judgment and surface inconsistencies a purely manual review might miss due to volume. The final decision still sits with a human analyst weighing the AI's output against context the model may not fully capture.

How Are Companies Building Sustainable AI Infrastructure at Scale?

Beyond perception management and investment screening, some companies are taking concrete steps to build AI infrastructure powered by renewable energy. Antrique, through its subsidiary Envolv Energy, has secured an initial 150 megawatts of long-term green energy capacity as a foundation for an integrated AI infrastructure platform in Brunei Darussalam.

This arrangement provides Antrique with long-term access to green energy for at least twenty years, with provisions for continued renewal beyond the initial term. This extended timeline provides visibility over future energy availability and economics, supporting infrastructure planning and capital deployment.

The strategy reflects a broader recognition that as AI applications advance and compute infrastructure evolves, access to scalable, reliable, and increasingly sustainable power is becoming a critical consideration for AI infrastructure development. Antrique is developing its energy ecosystem in parallel with its compute platform to support accelerating AI adoption and successive generations of AI infrastructure.

Building on the initial 150 megawatts, Antrique and Envolv intend to progressively develop a dedicated green energy ecosystem in Brunei with a long-term target of 1 gigawatt. The strategy is designed to provide a more autonomous and resilient energy foundation for large-scale AI infrastructure, reducing reliance on the national grid as the sole source of power while progressively integrating renewable generation, energy storage, and intelligent energy management.

Antrique is also actively engaging with the Massachusetts Institute of Technology on research and emerging applications in renewable energy, energy systems, and intelligent infrastructure, with a focus on identifying technologies and approaches that can support the platform's future development.

What Should Companies Do Now?

The gap between AI energy perception and reality creates both a challenge and an opportunity for companies. The challenge is managing consumer expectations and building trust despite widespread misconceptions. The opportunity is demonstrating genuine sustainability value by using AI to solve real environmental problems across operations, supply chains, and product development.

"The sustainable AI conversation needs to move beyond energy consumption to where AI can create tangible business and sustainability value. The companies leading today are using AI to improve operational and energy efficiency, sell better by turning sustainability into commercial advantage, and protect better by identifying and managing climate risks," said Wissam Yassine.

Wissam Yassine, Partner and Middle East Sustainability practice leader, Bain & Company

For investors, the lesson is that AI-powered tools can help identify genuinely sustainable companies, but only when combined with human judgment and awareness of the tools' limitations. For companies, the imperative is to demonstrate the value AI can create while managing its environmental footprint and wider risks responsibly. Those that get this balance right will be better positioned to scale AI and build trust with consumers, investors, and regulators.