Tech companies are combining traditional artificial intelligence with generative AI, while claiming the energy-guzzling technology could help prevent climate change. According to a report.
An analysis of 154 statements revealed that most claims that AI can help prevent climate change refer to machine learning, not the energy-hungry chatbots and image generation tools that are driving the explosive growth of gas-guzzling datacenters in the region.
Research, including non-profit organizations beyond fossil fuels And climate action against disinformationCouldn’t find a single example where popular tools like Google’s Gemini or Microsoft’s Copilot were leading to “material, verifiable and substantial” reductions in planet-warming emissions.
Energy analyst and report author Ketan Joshi said the industry’s strategy was “distracting” and relied on tried-and-tested methods that amounted to “greenwashing”.
He compared this to fossil fuel companies that advertise their modest investments in solar panels and exaggerate their carbon capture potential.
“These technologies save only a small fraction of emissions relative to the massive emissions of their core business,” Joshi said. “Big tech took that approach and advanced and expanded it.”
Most of the claims examined came from International Energy Agency (IEA) reports, which were reviewed by major tech companies, and corporate reports from Google and Microsoft.
According to the analysis, the IEA report – which devoted two chapters to the potential climate benefits of traditional AI – had an almost equal split between academic publications, corporate websites, and claims that had no evidence. For Google and Microsoft, most of the claims lacked evidence.
The analysis, released during the AI Impact Summit in Delhi this week, argues that the tech industry has misleadingly presented climate solutions and carbon pollution as a package deal by confusing types of AI.
Sasha Lucioni, AI and Climate Leadership hugging faceAn open-source AI platform and community, Joe, who was not involved in the report, said it added nuance to a debate that often lumps together very different applications.
“When we talk about AI that is relatively bad for the planet, it’s mostly generative AI and large language models,” said Lucioni, who has pushed the industry to be more transparent about its carbon footprint.
“When we talk about AI that is ‘good’ for the planet, it is often predictive models, extractive models, or old school AI models.”
The analysis found that even for traditional AI, Green’s claims rely on weak forms of evidence that were not independently verified. Only 26% of the green claims studied cited published academic research, while 36% provided no evidence.
One of the early examples identified in the report was the widespread claim that AI could help reduce 5-10% of global greenhouse gas emissions by 2030.
This figure, which Google repeated as recently as April last year, came from a report from BCG, a consulting firm, which cited a blogpost written in 2021 that attributed the figure to its “experience with customers.”
According to BloombergNEF, datacenters consume only 1% of the world’s electricity, but their share of US electricity is projected to more than double to 8.6% by 2035. The IEA estimates that they will account for at least 20% of the rich world’s electricity demand growth by the end of the decade.
While for large language models like ChatGPT the energy consumption of a simple text query can be as low as turning on a lightbulb for a minute, partial industry disclosures show, it increases significantly for complex tasks like video production and in-depth research, and some energy researchers have been intrigued by the speed and scale of its growth.
A Google spokesperson said: “Our projected emissions reductions are based on a robust validation process based on the best available science, and we have transparently shared the principles and methodology that guide it.”
Microsoft declined to comment, while the IEA did not respond to requests for comment.
Joshi said the discussion about the climate benefits of AI needs to be “brought back to reality”.
“The false coupling of a big problem and a small solution serves to distract attention from the very preventable harm being caused through unrestricted datacenter expansion,” he said.
