August M
A widely cited claim that artificial intelligence could help cut 5–10% of global greenhouse gas emissions by 2030 has become central to tech industry messaging around climate solutions. But reporting by The Guardian suggests that some of these optimistic projections may rest on limited evidence.
According to The Guardian, analysts reviewing public claims about AI’s environmental impact found that the foundation for certain high-profile estimates was weaker than often presented.
Origins of the 5–10% Emissions Claim
One frequently repeated estimate — that AI could reduce global emissions by as much as 5–10% within the decade — can be traced back to a consultancy report commissioned by Google. As The Guardian reported, that document cited a company blogpost referring to its “experience with clients” as part of the basis for the projection.
The newspaper noted that such figures have been echoed in corporate sustainability messaging, even though independent academic verification appears limited.
Review of 154 Public Statements
The scrutiny stems from an analysis of 154 public statements about AI and climate impact. The research was commissioned by nonprofits including Beyond Fossil Fuels and Climate Action Against Disinformation.
As first reported by The Guardian, the review found that companies often blur the distinction between traditional machine learning systems and newer generative AI tools when discussing environmental benefits.
The study did not identify any instance in which widely used generative systems — such as Google’s Gemini or Microsoft’s Copilot — produced what it described as a “material, verifiable, and substantial” reduction in greenhouse gas emissions.
Evidence Behind Environmental Claims
The analysis also examined the strength of evidence supporting corporate green claims. According to The Guardian’s reporting, only 26% of the environmental benefit statements referenced peer-reviewed academic research, while 36% included no supporting evidence at all.
Ketan Joshi, an energy analyst and author of the report cited by the newspaper, described industry messaging as “diversionary” and said it relied on tactics that amount to “greenwashing.”
“These technologies only avoid a minuscule fraction of emissions relative to the massive emissions of their core business,” Joshi said. “Big tech took that approach and upgraded and expanded it.”
Rising Energy Demand From Datacentres
The debate comes at a time when electricity consumption tied to AI infrastructure is increasing.
Datacentres currently account for about 1% of global electricity use. BloombergNEF projects that their share of US power consumption could rise to 8.6% by 2035. The International Energy Agency has said datacentres may drive at least 20% of electricity demand growth in advanced economies by the end of the decade.
While a single text query to a large language model may require relatively modest energy, more complex applications — including video generation and intensive computational research — demand substantially more processing power.
Energy analysts cited by The Guardian say this rapid expansion complicates arguments that AI’s climate benefits will outweigh its growing electricity footprint.
Industry Response
A Google spokesperson told The Guardian: “Our estimated emissions reductions are based on a robust substantiation process grounded in the best available science, and we have transparently shared the principles and methodology that guide it.”
Microsoft declined to comment, and the International Energy Agency did not respond to requests for comment, the newspaper reported.
Joshi said the broader conversation about AI’s climate role needs to be “brought back to reality,” warning that overstated environmental promises risk distracting from the impacts of expanding digital infrastructure.
Source: Reporting by The Guardian; BloombergNEF; International Energy Agency.
