Google has published a report outlining the energy consumption of its Gemini applications per query. The median energy used for a typical query is 0.24 watt-hours, comparable to the energy required to operate a standard microwave for approximately one second. Additionally, the report highlights average metrics for water usage, which amounts to five drops per query, as well as the associated carbon emissions linked to a text prompt processed by Gemini.
This report represents a significant level of transparency from a major technology company regarding the energy impact of its AI solutions. It also details the methodologies utilized to derive the reported estimates, providing insights into the calculations involved.
Earlier this year, a series published by MIT Technology Review focused on the relationship between artificial intelligence and energy consumption, noting that many major AI firms were reluctant to disclose their energy usage data per query. Google’s new findings address this gap, offering researchers and industry analysts a clearer understanding of the energy demands of AI applications.
In a separate context, an individual shared their experience of voluntarily submitting their DNA profile to a private genealogical database, FamilyTreeDNA, allowing law enforcement access to this information. This decision was influenced by a desire to challenge prevailing concerns regarding DNA privacy, especially in light of techniques like forensic investigative genetic genealogy (FIGG), which have been instrumental in resolving numerous criminal cases. This use of genetic data has sparked ongoing debates about privacy and the implications of sharing genetic information with law enforcement.
Source: https://www.technologyreview.com/2025/08/22/1122350/the-download-googles-ai-energy-expenditure-and-handing-over-dna-data-to-the-police/