At your next board meeting where AI appears on the agenda, add one question: Can our AI growth plans and our climate commitments succeed together? That simple litmus test will reveal whether you’re treating energy demand as a governance issue or leaving it as a technical detail, says Tim Weiss, CEO and co-founder of Optera. The real risk may not be AI itself but letting AI decisions and climate strategy live in separate parts of the organization.
Compliance and risk leaders are seeing a flood of headlines warning that AI will overwhelm power grids, as the data center construction boom continues. Behind those headlines are real concerns about data center growth, climate targets and community impacts. The challenge is to separate credible risk from speculation so boards and executives can govern AI growth without losing sight of climate commitments.
The latest projections from the International Energy Agency suggest that global electricity use from data centers could more than double by 2030, reaching around 945 terawatt hours and representing just under 3% of global power demand, with AI as a major driver.
Goldman Sachs research forecasts that global power demand from data centers could increase 50% by 2027 and by as much as 165% by 2030 compared with 2023. These figures belong in boardrooms and risk assessments. They also come with a level of uncertainty that governance teams need to understand.
AI demand is exposing how slow and carbon-intensive traditional power development can be. That mismatch between rapid load growth and long build times is already steering utilities and investors toward cleaner projects that can be built and scaled quickly. For organizations that care about carbon management and regulatory credibility, the question is not if AI appears in the risk register but whether AI growth and climate strategy are being managed in tandem.
The numbers matter, but so do their limits
The headline numbers on AI and energy are attention-grabbing for a reason. Growing AI adoption requires more computing power, and that power has to come from somewhere. According to the IEA estimates, data center electricity consumption is growing more than four times faster than total electricity consumption from other sectors, while the European Central Bank similarly warns that AI-related data center loads could influence commodity prices if growth is left unchecked.
Several underlying assumptions are still in flux, including how quickly models will become more efficient and which AI workloads will actually prove valuable enough to justify their energy use. In other words, today’s models are early versions. Over time, they are likely to become more efficient and more targeted to specific tasks.
Hardware is improving, model design is evolving, and many organizations are still in the early stages of figuring out which AI use cases truly add value. The first phase of any technology wave tends to reward experimentation. Over time, higher energy bills and stricter disclosure will push organizations to tighten their focus.
The key is to treat forecasts as scenario inputs rather than fixed outcomes. The range between conservative and aggressive projections is wide. Boards need to see that range clearly and know which assumptions management is relying on.
Fractured & Fraught — but Still Potentially Profitable: The State of ESG in 2025
As we approach the end of 2025, the state of ESG is in flux: Trump Administration executive orders have sought to roll back environmental regulations, while the EU has softened its landmark ESG disclosure requirements and California has pressed on with its similar regulations. As
Read moreDetailsIncreasing power demand is accelerating renewable power
Beneath the AI headlines, the energy system itself is changing quickly. Renewable power has become the main source of new capacity worldwide. Recent IRENA data shows that in 2024, renewable power represented more than 90% of all new power capacity. Almost all of that growth came from solar and wind projects.
This surge reflects renewables being the lowest cost option for new power generation, and they can be built much faster than large coal or gas plants. For utilities and developers trying to respond to AI-driven load requests on tight timelines, those economics and construction realities matter more than any single climate pledge.
Corporate buyers are reinforcing this trend. Clean energy procurement by corporations increased 29% in 2024, led by data center operators and their customers. The private sector has become a central driver of budding renewable projects as organizations look for ways to support growing digital operations while meeting climate expectations set by the European Union and California.
For governance teams, these contracts go beyond sustainability anecdotes and act as long-term financial commitments that interact with climate targets, disclosure obligations and future regulatory changes. They should be treated with the same rigor as any other major supply agreement or infrastructure investment.
A new lens on carbon management and exposure
AI energy demand will not be the single largest source of emissions that compliance teams oversee. Manufacturing, logistics and supply chain emissions will continue to dominate most carbon profiles. What makes AI interesting is that its emissions are concentrated in highly visible, auditable nodes.
Data centers are fixed assets with measurable electricity use and clear owners. That makes AI-related energy use easier to manage than other sources of emissions, which could be spread across thousands of suppliers and downstream partners. When a company signs a contract that triggers a new data center build or expansion, it is possible to trace a specific piece of the emissions profile back to that decision. The relative clarity of data center emissions can draw attention away from larger but less visible sources of Scope 3 emissions if governance is not careful about proportionality.
Organizations can extend existing governance processes to cover AI-related emissions, as these already cover energy and infrastructure issues.
Governance moves for the next 12 to 24 months
The organizations that navigate AI power demand most effectively will treat it as part of their broader climate and risk strategy, not as a separate technology issue. Two practical steps can help.
Strengthen diligence and contract terms with data center and energy partners
When reviewing or renewing contracts with data center providers and energy suppliers, expand due diligence questions to cover energy sourcing, renewable procurement plans, expected location of new builds and water use practices where relevant.
Integrate these questions into existing third-party risk and supplier engagement processes. Where appropriate, consider contract language that requires notice of major changes to energy sourcing or siting that could affect emissions or local community impacts.
Align AI decisions with climate commitments and risk appetite
Ensure that climate and risk stakeholders have a voice when large AI investments come forward. Carbon accounting teams can help clarify how new loads will show up in Scope 2 and Scope 3 emissions and whether current transition plans can accommodate them. Risk and compliance leaders can frame AI-related energy exposure alongside other transition risks, such as future carbon prices, local disclosure rules and the potential for policy shifts.
These steps do not require perfect data or long-range certainty. They require a structured way to bring AI and energy considerations into decisions that organizations are already making.
AI and climate commitments belong in the same conversation
The transition to a low-carbon economy is already underway, driven by a mix of regulation, market pressure and physical climate risk. That trajectory is unlikely to reverse, even if individual policies shift from year to year.
If AI-driven power growth is managed in isolation, organizations risk undermining their own climate targets and exposing gaps in governance. The remedy is to integrate AI decisions into carbon management, carbon accounting and broader risk frameworks. The same pressure that currently worries boards can help accelerate cleaner energy deployment and more transparent reporting.
The litmus test is simple. At the next board, audit or risk committee meeting where AI appears on the agenda, add one question: How confident are we that our AI growth plans and our climate commitments can succeed together? That conversation will reveal quickly whether AI energy demand is being treated as a central governance issue or left as a technical detail.
The sooner organizations bring those threads together, the better positioned they will be for whatever mix of AI adoption, regulation and climate risk arrives over the rest of this decade.
Tim Weiss is co-founder and CEO of Optera, which offers enterprise emissions management software. In that role, he oversees the company’s operations, product and business strategy. Tim has been at the forefront of corporate sustainability for nearly a decade, supporting Fortune 100 companies and leading non-governmental organizations like the World Economic Forum and CDP in advancing science-based targets, climate strategy and emissions accounting. 
