Share
Tech giants are facing mounting accusations of “greenwashing” as a new report claims they are misleading the public by conflating traditional machine learning with energy-intensive Generative AI.
The research, commissioned by nonprofits including Climate Action Against Disinformation, suggests that the industry is using “diversionary tactics” to mask the massive environmental cost of the current AI gold rush.
According to energy analyst Ketan Joshi, the report’s author, tech companies frequently cite the climate benefits of “old-school” predictive models – which can optimize power grids or track deforestation – to justify the explosive growth of gas-guzzling data centres required for generative tools such as OpenAI’s ChatGPT or Microsoft’s Copilot.
The analysis of 154 corporate and industry statements failed to find a single example where Generative AI led to a “material, verifiable, and substantial” reduction in global emissions.
The debate centres on the stark difference in energy profiles. While predictive AI uses relatively modest resources, Generative AI requires massive clusters of high-performance GPUs. Sasha Luccioni, a climate lead at Hugging Face, notes that when the industry discusses AI that is “bad for the planet,” it is almost exclusively referring to large language models and image generators.
This surge in demand has sparked a critical question. Can Generative AI ever be carbon neutral? Certainly, the hardware fuelling this revolution generates immense heat, requiring sophisticated cooling systems that often consume vast amounts of water and electricity.
While companies like Google claim their emission reduction estimates are based on robust science, data centres are projected to account for 20% of electricity demand growth in wealthy nations by the end of the decade.
The report likens these tech claims to fossil fuel companies overstating the potential of carbon capture while their core business continues to drive pollution. As complex functions such as video generation and deep research proliferate, analysts argue that the narrative of AI as a climate saviour is being used to distract from the “preventable harms” of unrestricted data centre expansion.
Unless transparency regarding the carbon footprint of GPUs and cooling improves, the industry’s green claims will remain under intense scrutiny.
Related Posts
Discover more from Tech Digest
Subscribe to get the latest posts sent to your email.