A company deployed AI across operations in 2025.
Customer service, HR, marketing, sales, everything got AI tools.
Energy consumption increased 300%. The company didn’t track it because nobody connected AI deployment to energy use.
Now it’s Q1 2026. ESG reporting season.
An activist investor asks during the earnings call: “How much energy do your AI systems consume?”
The company doesn’t know. Their sustainability report doesn’t mention AI. Their carbon reduction targets suddenly look impossible.
And they realize every AI query uses significantly more energy than traditional computing.
Why AI Uses So Much Energy
AI models require massive computational power.
Training models: Training large AI models consumes enormous energy. GPT-4 level models use megawatt-hours of electricity.
Running inference: Every AI query (chatbot response, image generation, analysis) requires significant computation compared to traditional software.
Data centers: AI requires specialized GPU infrastructure that uses more energy than traditional servers.
Cooling requirements: AI data centers generate massive heat, requiring additional energy for cooling.
A single AI query can use 10-100x more energy than a traditional search query.
When companies deploy AI at scale, energy consumption increases dramatically.
The ESG Reporting Gap
Most companies report on energy consumption and carbon emissions in ESG reports.
But they’re not measuring AI’s energy impact separately, don’t know how much energy AI systems consume, can’t report AI carbon footprint to investors, and haven’t updated sustainability targets to account for AI.
ESG reports show declining or flat energy use, while actual consumption increased due to unreported AI systems.
This creates a reporting gap that investors are starting to question.
SEC Climate Disclosure Rules
SEC climate disclosure rules (finalized 2024, phased implementation) require companies to disclose Scope 1 and Scope 2 emissions, with some requirements for material climate risks.
AI energy consumption affects these disclosures by increasing Scope 2 emissions (purchased electricity), potentially creating material climate risks if AI energy use is substantial, and requiring disclosure if AI affects the ability to meet climate targets.
Companies deploying significant AI without measuring energy impact may be under-reporting climate-related information.
Investor Scrutiny
Investors are asking specific questions about AI and sustainability.
Earnings calls: “How much energy do your AI systems consume?”
ESG surveys: “What is the carbon footprint of your AI infrastructure?”
Shareholder proposals: Some investors are filing proposals requiring AI energy disclosure.
ESG ratings: Rating agencies are starting to evaluate AI energy impact in sustainability scores.
Companies that can’t answer these questions look unprepared or like they’re hiding negative information.
The Greenwashing Risk
Companies make sustainability commitments like carbon neutrality by 2030, renewable energy targets, or emissions reductions.
Then they deploy AI that triples energy consumption.
If sustainability reports don’t mention AI’s energy impact, it looks like greenwashing, making environmental claims while ignoring factors that undermine those claims.
Regulators and investors treat greenwashing seriously. It creates reputational and regulatory risk.
What Companies Should Measure
To accurately report AI’s environmental impact, companies need to track AI energy consumption by system, carbon emissions from AI operations, comparison of AI energy use to traditional systems, and progress toward energy efficiency in AI deployment.
Most companies have no systems to measure any of this because they deployed AI without considering the energy impact.
What Should Happen Now
Measure AI energy consumption: Implement tracking for energy used by AI systems separately from general IT infrastructure.
Update ESG reporting: Include AI energy impact in sustainability reports and climate disclosures.
Revise sustainability targets: Adjust carbon reduction goals to account for AI energy consumption or commit to offsetting AI emissions.
Evaluate AI efficiency: Choose AI systems and deployment methods that minimize energy use where possible.
Prepare for investor questions: Have data ready when investors ask about AI sustainability impact.
Communicate proactively: Explain AI’s energy impact and what the company is doing to manage it before investors discover the reporting gap.
Brian Alexander
Founding Partner, VisionAI+ Consulting Group
Former General Counsel, AI Engineering Firm
Certified ISO 42001 Lead Auditor