A new report commissioned by seven environmental organizations found that 74% of big tech companies' claims about AI reducing greenhouse gas emissions lack credible evidence, with many based on little more than blog posts rather than rigorous analysis. The findings challenge the narrative that artificial intelligence will be a net climate benefit, suggesting instead that tech firms are using sophisticated greenwashing tactics to obscure the environmental costs of powering massive data centers.

What's Behind Big Tech's Climate Claims?

Google's Chief Sustainability Officer, Kate Brandt, claimed in a co-authored op-ed that AI could prevent 5 to 10 percent of global emissions by 2030, equivalent to the entire European Union's annual emissions . The claim was based on a paper jointly published by Google and consulting firm BCG. However, independent climate and energy analyst Ketan Joshi, who authored a report titled "The AI Climate Hoax" commissioned by environmental organizations, discovered that Google and BCG's analysis was essentially a blog post rather than peer-reviewed research .

Google also touts its use of AI to lower emissions in transportation and energy sectors. The company claimed that incorporating AI in five of its products enabled others to collectively reduce emissions by an estimated 26 million metric tons in 2024 . Yet when examined closely, these figures often misattribute emissions reductions to AI tools when they actually result from the underlying technology itself. For example, Google's Solar API tool received credit for emissions avoided by rooftop solar installations, even though the mapping tool did not cause the installations to exist .

How Are Tech Companies Misleading the Public?

Joshi's analysis of 154 statements by tech companies found that only 26% cited published academic papers as evidence, while 36% provided no evidence whatsoever . The report identified five specific greenwashing tactics that tech firms employ, mirroring the United Nations definition of misleading the public about environmental protection efforts:

• Hiding Emissions: Companies report lower emissions through accounting strategies like purchasing renewable energy certificates, despite fossil fuels actually powering their operations.
• Empty Promises: Tech firms make vague pledges to deploy futuristic technologies such as nuclear fusion, carbon capture and storage, or space-based solar power, which are "incredibly expensive and unlikely to work."
• Downplaying: Companies share small, relatively meaningless disclosures about AI's sustainability, such as reporting emissions from a single chatbot query while omitting the entire system's emissions.
• Tactical Fatalism: Industry leaders label climate goals as unachievable, with Google's former CEO Eric Schmidt stating in 2024 that "we're not going to hit the climate goals anyway."
• Empty Ambition: Some companies pair ambitious climate targets with notably insufficient action to achieve them.

"The promises of planet-saving tech remain hollow, while AI data centres breathe life into coal and gas every day. These claims of climate benefit are unjustified and overhyped, and could cover up irreversible damage being done to communities and society," said Ketan Joshi.

Ketan Joshi, Independent Climate and Energy Analyst

Joshi noted that "the novelty of generative AI and data centre expansion has lent itself to a novelty of greenwashing tactics" . Consumer-facing generative AI systems like ChatGPT have not achieved any "material, verifiable, and substantial level of emissions reductions," according to the report .

Joshi

What Are the Real Climate Costs?

While tech companies make optimistic claims about AI's climate benefits, global temperatures continue rising. The average global temperature increase in 2025 was 1.44 degrees Celsius, just below the 1.5-degree threshold set by the Paris Agreement . The World Meteorological Organization called 2025 "one of the warmest years on record" .

Beyond carbon emissions, researchers at Harvard T.H. Chan School of Public Health are documenting serious health costs from AI data centers powered by on-site fossil fuel plants. Michael Cork, a postdoctoral researcher in biostatistics, has analyzed air pollution impacts from multiple proposed and existing data center facilities . His research reveals that fine particulate matter from these power sources drives nearly 90% of health impacts from air pollution, yet these costs are often excluded from infrastructure planning decisions .

In one high-profile case, Cork and colleagues estimated that the Vantage Data Center in Loudoun County, Virginia, which uses on-site gas turbines for power, generates between $53 million and $99 million in annual health damages from associated air pollution . The damages are driven primarily by an estimated 3.4 to 6.5 additional premature deaths per year across the impacted region, along with increased hospital admissions, asthma-related outcomes, and lost productivity .

How Can Communities Protect Themselves From Data Center Pollution?

Experts emphasize that closing the gap between permitting decisions and health impact analysis is essential. Cork explained that independent, science-based health impact analysis can inform regulatory decisions at critical moments in the permitting process . In Fluvanna County, Virginia, for example, health impact analysis shaped public discussion leading up to a Planning Commission vote to recommend denial of a gas plant expansion .

Communities and regulators can take several steps to ensure data center projects account for public health costs:

• Demand Independent Analysis: Require third-party health impact assessments before approving data center projects, rather than relying solely on company-provided environmental reports.
• Account for Air Pollution: Ensure permitting decisions include comprehensive analysis of fine particulate matter and other air pollutants, not just carbon emissions or grid capacity concerns.
• Prioritize Renewable Energy: Mandate that data centers use renewable energy sources rather than on-site fossil fuel power plants, eliminating local air pollution risks.
• Engage Vulnerable Communities: Ensure that residents in high-risk areas, including children, the elderly, and people with pre-existing health conditions, have a voice in permitting decisions.

"We can't solve a problem we haven't measured. Right now, permitting decisions for data centers are being made without any clear accounting of their public health costs, and that gap means health damages are systematically left out of these important decisions," explained Michael Cork.

Michael Cork, Postdoctoral Researcher in Biostatistics, Harvard T.H. Chan School of Public Health

The disconnect between tech industry climate claims and measurable environmental outcomes reflects a broader pattern. While some companies are developing genuine solutions, such as Woodchuck, an AI-powered startup that diverts wood waste from construction sites into renewable energy, the majority of big tech's public statements about AI and climate appear designed to manage perception rather than drive real change . Woodchuck diverted more than 13,000 tons of wood waste from landfills in 2025 alone, converting that material into approximately 905 million BTUs of renewable energy, demonstrating that measurable climate impact is possible when economic incentives align with environmental outcomes .

As global temperatures approach critical thresholds, the gap between tech industry rhetoric and reality has never been more consequential. Experts argue that it is time for companies like Google and Microsoft to move beyond aspirational statements and take concrete steps to reduce emissions, ideally before the world exceeds the 1.5-degree warming limit .