AI's Climate Benefits Overstated as 'Greenwashing', Energy Analyst Warns
Discourse surrounding artificial intelligence's potential to combat climate change must be brought back to reality, according to energy analyst Ketan Joshi. In a damning report, claims that AI can help fix the climate crisis are dismissed as diversionary tactics and outright greenwashing by the tech industry.
Industry Conflates AI Types to Mislead on Environmental Impact
The analysis, commissioned by nonprofits including Beyond Fossil Fuels and Climate Action Against Disinformation, scrutinized 154 statements from tech giants and found a troubling pattern. Most assertions that AI could avert climate breakdown refer to traditional machine learning, not the energy-intensive generative AI tools like chatbots and image generators that are driving explosive growth in gas-guzzling datacentres.
Joshi, author of the report, likened this approach to fossil fuel companies overstating minor investments in renewables. "These technologies only avoid a minuscule fraction of emissions relative to the massive emissions of their core business," he stated. "Big tech took that approach and upgraded and expanded it."
Lack of Evidence and Misleading Claims Exposed
The research did not uncover a single instance where popular tools such as Google's Gemini or Microsoft's Copilot led to a material, verifiable, and substantial reduction in planet-heating emissions. Key findings include:
- Most claims originated from an International Energy Agency report reviewed by tech companies, with roughly half lacking evidence.
- For Google and Microsoft, the majority of climate benefit assertions had no supporting proof.
- Only 26% of green claims cited published academic research, while 36% provided no evidence at all.
Sasha Luccioni, AI and climate lead at Hugging Face, emphasized the need for nuance, noting that "when we talk about AI that's relatively bad for the planet, it's mostly generative AI and large language models." In contrast, beneficial applications often involve older, predictive models.
Energy Consumption Surges Amidst Unverified Promises
While datacentres currently consume about 1% of global electricity, projections indicate a sharp rise. In the US, their share is expected to more than double to 8.6% by 2035, and the IEA predicts they will account for at least 20% of electricity demand growth in wealthy nations this decade.
Although a simple text query to a model like ChatGPT might use as little energy as running a lightbulb for a minute, complex functions such as video generation and deep research drive significantly higher consumption. This rapid expansion has alarmed energy researchers.
Tech Responses and Calls for Transparency
Google defended its position, stating its emissions reductions are based on "a robust substantiation process grounded in the best available science." Microsoft declined to comment, and the IEA did not respond to requests.
Joshi concluded that the false coupling of major climate problems with minimal solutions distracts from preventable harms caused by unchecked datacentre growth. He urged a return to factual discourse to address the real environmental impacts of AI's energy-hungry infrastructure.
