The increasing use of AI, such as ChatGPT, comes at a significant environmental cost. AI-powered chatbots, for example, consume at least 10 times more electricity than a standard Google search. If all Google searches implemented generative AI, the energy consumption could be equivalent to that of a country like Ireland. The surge in energy demand due to AI technologies will lead to increased carbon emissions and a greater reliance on fossil fuels. Additionally, AI has a substantial water footprint, with ChatGPT using as much as a 16-ounce bottle of water in just 10 queries. Concerns over the environmental impact of AI have arisen globally, as it could hinder the adoption of green energy, result in higher electricity bills, and raise the risk of blackouts. Experts advocate for transparency from tech companies regarding the power and water consumption of AI applications, suggesting the provision of information to users to make informed decisions. Data centers, housing the servers that support the internet, have always been significant energy users, but the specialized computer chips required for generative AI consume even more energy and require additional cooling.

Despite the unknowns regarding the benefits and risks of AI, companies are increasingly integrating it into their products. Google, for instance, introduced "AI Overviews" on its search engine, generating answers from search results, though not always accurately. Users currently lack the option to opt out of this feature or avoid the increased energy and water usage. Google's commitment to achieving "[net-zero](https://en. wikipedia. org/wiki/Net-zero_emissions)" emissions by 2030 does not seem to be progressing, as its carbon emissions have increased by 13% since 2023, with a 48% increase since 2019. The company expects emissions to continue rising before eventually decreasing. Google's data centers consumed 6. 1 billion gallons of water in 2023, a 17% increase from the previous year. While Google has acknowledged the need to address AI's environmental impact, it has not disclosed the specific power usage attributable to AI in its reports, preventing accurate calculations of the additional electricity consumption resulting from AI Overviews.