Artificial intelligence (AI) models have been under scrutiny over their energy consumption, with OpenAI CEO Sam Altman trying to ease concerns by comparing it to human development. He said that training an AI model takes about 20 years of life and all the food consumed during that time to become smart.
However, experts disagree, citing datacenter energy consumption’s significant impact on the environment. Datacenters now account for around 1.5% of global electricity consumption, which is projected to increase by over 15% each year until 2030. Environmental groups have called for a moratorium on building new datacenters in the US, arguing that they are disrupting communities and threatening Americans’ economic, environmental, climate, and water security.
AI advocates like Altman downplay concerns about datacenter energy consumption, but some experts dispute their claims. Mike Weinstein, director of sustainability at Southern New Hampshire University, is skeptical of the argument that the power required by AI infrastructure justifies its benefits. He questions whether technology will help alleviate global problems, citing a recent report from OpenAI that found 70% of messages to ChatGPT were not work-related.
Altman’s remarks sparked backlash online, with some describing them as dystopian and comparing his comments to Black Mirror’s exploration of potential harms from new technology. As the debate continues, experts emphasize the need for sustainable energy solutions to support growing AI infrastructure.
Source: https://www.theguardian.com/technology/2026/feb/23/sam-altman-openai-energy-use-datacenters