In the modern digital age, generative artificial intelligence (AI) is becoming increasingly popular, with one in three French people reportedly having consulted an AI for internet searches. While AI provides immediate answers, the energy costs associated with these responses are often overlooked. A simple query can consume 0.363 watt-hours (Wh) of electricity, nearly 2% of a smartphone’s full charge.
The Growing Energy Appetite of AI
Generative AI models, while efficient in providing quick responses, require substantial energy, especially as their tasks grow more complex. From model development to everyday use, these AI systems are notably energy-intensive. Recent developments highlight the burgeoning energy demands of AI technologies.
In March, OpenAI’s CEO, Sam Altman, acknowledged the energy challenges faced by the company. Following the launch of a new image generation feature in ChatGPT, the demand surged. Users eagerly tested the AI’s ability to transform photos into artistic styles, particularly those reminiscent of Studio Ghibli, a renowned Japanese animation studio.
Within days, ChatGPT experienced overwhelming user traffic, with one million new users signing up within an hour. This surge threatened service performance, underlining the significant energy consumption associated with AI operations.
Projections and Energy Impact
The explosive growth of AI is projected to more than double the global electricity demand from data centers by 2030. According to an International Energy Agency (IEA) report, this demand could reach approximately 945 terawatt-hours (TWh), surpassing Japan’s current total electricity consumption. By 2030, data centers are expected to consume nearly 3% of the world’s electricity.
In the United States, data centers are anticipated to contribute significantly to the increase in electricity demand. The IEA notes that power consumption by data centers could account for almost half of the projected growth in electricity demand over the next decade.
Initiatives and Innovations
To address the energy challenge, several initiatives are underway. Former President Donald Trump announced the Stargate project, backed by a $500 billion package of public and private funds, aimed at constructing up to 10 new data centers. In a bid to connect data centers directly to cleaner energy sources, companies like Meta and Microsoft are planning partnerships with nuclear power plants.
These efforts are predicated on the expectation that AI technology will achieve substantial success in the coming years. As AI models become integral to platforms like Bing, WhatsApp, and Google, their inclusion in various apps on smartphones is expected to raise questions about their environmental impact.
The Future of AI and Energy
The integration of AI into everyday applications highlights the need for sustainable energy solutions. As AI technologies continue to evolve, balancing their benefits with environmental considerations will be crucial. Companies and governments must collaborate to develop energy-efficient practices that support the growth of AI without compromising ecological integrity.
The ongoing advancements in AI present both opportunities and challenges. While AI offers transformative potential across industries, its energy demands necessitate innovative approaches to minimize environmental impact.
Note: This article is inspired by content from https://www.lemonde.fr/en/les-decodeurs/article/2025/06/14/artificial-intelligence-consumes-massive-amounts-of-energy-here-s-why_6742347_8.html. It has been rephrased for originality. Images are credited to the original source.