The MIT Technology Review piece argues that the climate case for AI has not yet been made. It says industries are not doing nearly enough to use AI to cut greenhouse-gas emissions, despite optimistic claims from AI companies. OpenAI CEO Sam Altman said in 2024 that AI would bring "astounding triumphs," including "fixing the climate." A 2023 op-ed by Google's head of sustainability and a coauthor said AI could reduce global emissions 5% to 10% by 2030.
The article cites a peer-reviewed study led by researchers at the London School of Economics' Grantham Research Institute and the firm Systemiq. That study estimated AI could cut global emissions by 3.2 to 5.4 billion tons of carbon dioxide annually by 2035, after subtracting AI's own energy footprint. The gains are concentrated in power, meat and dairy, and vehicles.
The article's AI-assisted review of public information found few companies pursuing the specific routes behind those projections. In the power sector, the study estimated AI-driven forecasting could raise renewable output from existing solar and wind facilities by up to 20%. Of the grid operators reviewed in China, the United States, and Europe, the article says only one, the UK's National Grid ESO, has a known working program of this kind. Its partnership with Open Climate Fix halved its solar-forecast error.
In meat and dairy, the article says NotCo and Kraft Heinz formed a joint venture using NotCo's Giuseppe platform to replicate animal-product taste and texture. It says few other alternative-protein companies use AI so deeply in product development. It also reports that venture capital into cultivated meat fell from a $989 million peak in 2021 to about $83 million last year.
In transportation, the article cites battery maker CATL, which it says controls roughly 40% of the global EV battery market. CATL reported launching an AI system that cut the time to screen candidate materials by two orders of magnitude, according to the company. Independent researchers found that Didi Chuxing cut per-trip emissions 16% on average using AI-optimized routing.
The article also warns that oil and gas companies are interested in AI partly to increase fossil fuel production. It adds that AI does little to shorten certification, permitting, and construction timelines.