The growth of computational load from AI increases the energy consumption of data centers. According to the International Energy Agency, their consumption could nearly double from 485 TWh in 2025 to 950 TWh by 2030. Against this backdrop, services are emerging that aim to offset the impact of AI with ecological projects. One of them is EcoGPT, linking the number of queries to funding for tree planting.
In the free version of the service, one tree is funded for every 200 generations, while for Pro subscribers, this figure is 100 queries. EcoGPT collaborates with Trees for the Future, and in September, the organization issued a certificate for 100,000 trees to the company. However, the service does not disclose detailed methodology for calculating the ecological effect and the share of revenue directed towards reforestation. Experts also note that the energy costs of AI operation cannot be directly compared to the number of trees planted: CO₂ absorption occurs gradually, and the survival rate of saplings depends on regional conditions and care quality.
The ecological model also depends on infrastructure. EcoGPT claims to use small models and hosts its main infrastructure on Google Cloud, but in certain scenarios, it may route queries through OpenRouter to third-party providers. Therefore, the complete energy and water footprint of individual operations cannot be publicly determined. Experts surveyed believe that transparent calculations, independent verification, and monitoring of results are necessary to substantiate climate claims. Without this data, tree planting alone does not allow for establishing what volume of emissions is actually offset.
Related articles
