The tech industry's claims that artificial intelligence can help mitigate climate change have been met with skepticism by experts, who argue that these claims are often based on flawed assumptions and a lack of evidence. According to a recent report, the industry's tactics are "diversionary" and rely on "greenwashing" to create a positive public image.
The report, commissioned by nonprofits including Beyond Fossil Fuels and Climate Action Against Disinformation, analyzed 154 statements from tech companies and found that most claims about AI's potential to reduce emissions were based on traditional machine learning, rather than the energy-hungry generative AI that is driving the sector's growth. Ketan Joshi, an energy analyst and author of the report, said that the industry's approach was similar to that of fossil fuel companies, which often advertise their modest investments in renewable energy while downplaying their overall emissions.
The analysis found that only 26% of the green claims studied cited published academic research, while 36% did not cite evidence at all. One of the earliest examples identified in the report was a claim that AI could help mitigate 5-10% of global greenhouse gas emissions by 2030, which was attributed to a blog post written by a consulting firm in 2021. As The Guardian has previously reported, the tech industry's claims about AI's potential to reduce emissions have been widely publicized, but often lack concrete evidence.
Greenwashing in the Tech Industry
Sasha Luccioni, AI and climate lead at Hugging Face, an open-source AI platform and community, said that the report added nuance to a debate that often lumped very different applications of AI together. "When we talk about AI that's relatively bad for the planet, it's mostly generative AI and large language models," Luccioni said. "When we talk about AI that's 'good' for the planet, it's often predictive models, extractive models, or old-school AI models."
The report's findings have significant implications for the tech industry, which is increasingly reliant on energy-hungry data centers to power its operations. Data centers consume just 1% of the world's electricity, but their share of US electricity is projected to more than double to 8.6% by 2035, according to BloombergNEF. The International Energy Agency (IEA) predicts that data centers will account for at least 20% of the rich world's growth in electricity demand to the end of the decade.
The Energy Consumption of AI
The energy consumption of AI is a complex issue, and the report found that while simple text queries to large language models may have a relatively low energy cost, more complex functions such as video generation and deep research require significantly more energy. A spokesperson for Google said that the company's estimated emissions reductions were based on a robust substantiation process grounded in the best available science, but the report's authors argue that the industry's claims about AI's potential to reduce emissions are often based on flawed assumptions and a lack of evidence.
Microsoft declined to comment on the report, while the IEA did not respond to requests for comment. Joshi said that the discourse around AI's climate benefits needed to be "brought back to reality," and that the false coupling of a big problem and a small solution served as a distraction from the very preventable harms being done through unrestricted data center expansion.
Expert Analysis and Recommendations
Experts argue that the tech industry needs to be more transparent about its carbon footprint and the energy consumption of its operations. Luccioni has pushed the industry to be more transparent about its carbon footprint, and the report's authors recommend that companies provide more detailed information about their energy consumption and emissions reductions. As the tech industry continues to grow and expand, it is essential that it prioritizes sustainability and reduces its environmental impact.
