This article first appeared on Diligent Market Intelligence's Voting newswire. To register for a demonstration and trial of the product, click here.
Investors are pressing Big Tech over the environmental costs of the AI build-out, particularly the growing energy and water demands of data centers.
Diligent Market Intelligence identified 14 AI-related shareholder proposals to face a vote at U.S.-based companies in the first half of 2026, nearly half of which raised concerns about data-center resource use.
"We are at the beginning of this AI boom," said Kelly Poole, climate and energy coordinator at As You Sow. "Investors see the opportunity in intervening early so that we can have conversations with companies about sustainably developing their AI infrastructure before we find ourselves further along when more infrastructure is built, where money has been spent, and it's harder to undo the environmental and social impacts."
Big Five pressure
Major tech companies bore the brunt of investor concern over AI during the recent proxy season, with four of the Big Five tech companies facing a push for greater accountability for the environmental and social risks linked to AI infrastructure.
At Amazon, investors sought disclosure on how surging AI and data-center energy demand could affect emissions commitments and asked the board to revisit costs as infrastructure expands. Meta Platforms faced similar pressure to explain how AI-driven data-center growth and its rising energy and emissions footprint fit with the company’s climate goals. At Alphabet’s June 5 annual meeting, scrutiny extended from power demand to the water needed to cool AI infrastructure. Microsoft's last AGM brought the issue into human-rights territory, with shareholders seeking an assessment of the risks of locating data centers in sensitive jurisdictions and the measures needed to mitigate them.
As You Sow co-filed the proposal targeting data center energy demand at Amazon and secured 18% backing from investors. As You Sow's Poole told DMI that data centers are now of record-breaking size, sometimes using the same amount of electricity and water as a major city.
"As the AI boom really took off two to three years ago, this became an infrastructure challenge as well. It shifted to talking about energy and water as major tech companies expand their data center footprint very rapidly and at a scale that is now impacting the resource shares that were available," Poole explained.
For Danielle Fugere, president of As You Sow, this season's AI proposals are a broader reflection of public sentiment towards large companies expanding their growing AI demands. "I think that this pushback against data centers is another sign of citizens saying we don't need billionaires to force these technologies and their data centers down into our communities and impact us."
Risks arising from the growing demand for data centers have also been flagged by others in a new season of engagement.
Trillium Asset Management has made the issue a core focus. "The purpose is to raise awareness among AI users, and in this case, we’re looking at companies that likely have a significant AI footprint, to link data center energy consumption to individual company carbon footprints," Andrea Ranger, director of shareholder advocacy at Trillium, told DMI. "With data center energy, water, noise, and negative visual impacts in the headlines, investors, regulators, and lawmakers are increasingly sensitive to data centers’ adverse outcomes."
The concern is also spreading beyond Big Tech. "I think we’re going to see AI proposals and engagements across a wide range of sectors, such as healthcare, financial services, insurance, consumer goods and retail. We’re going to see this concern more broadly everywhere," said Madison Kreiger, activism and impact research analyst at NorthStar Asset Management.
NorthStar was one to move beyond tech to file at Digital Realty Trust and called for reporting on region-level water metrics on its exposure to water-related risks, claiming that the real estate company does not factor future AI growth in the tracking of water-based cooling in its operations. Following the vote, the advocacy group warned that water risk at data centers is now a portfolio risk.
"One of the main driving forces behind investors’ concerns is the pace of AI deployment outrunning the pace of governance," said Kreiger. "Investors rely on policy, but when policy doesn’t come through fast enough, it comes down to the companies themselves to put those guardrails in place."
A defining test
Ceres describes the AI build-out as a defining test for the U.S. electric power sector over the next three to five years. In its investor guidance, the advocacy group says data centers currently use 4% to 5% of U.S. electricity, with demand expected to rise rapidly over the next decade.
"As data centers become more prevalent around the country, community impacts (land use, pollution, water stress, electricity costs) are becoming more obvious," the report finds. "The AI data center boom has arrived at a moment when the U.S. electric grid is simultaneously under pressure to become cleaner, modernize, and expand at a pace it has not attempted in generations."
