Software Developers on Climate Change, AI, and Sustainable Software: A survey of GitHub users, Summer 2026


4. AI and Environmental Impact

4.1 A majority of GitHub users say AI systems contribute to climate change.

About six in ten respondents (63%) say AI systems (such as training and running large language models) are contributing to climate change either “a great deal” (28%) or “a moderate amount” (34%). One in three (33%) say AI systems are contributing “only a little” (21%) or “not at all” (12%). Few (5%) say they don’t know.This bar chart shows the percentage of GitHub users who believe that AI systems (e.g., training and running large language models) are contributing to climate change a great deal, a moderate amount, only a little, not at all, or say they don't know. A majority of GitHub users say AI systems contribute to climate change. Data: github, Spring 2026. Refer to the data tables in Appendix 1 of the report for all percentages.4.2 About seven in ten GitHub users are concerned about the environmental impact of AI systems.

About seven in ten respondents (71%) say they are either “very concerned” (39%) or “somewhat concerned” (32%) about the environmental impact of AI systems, including the energy and water use and carbon emissions associated with training and running large language models. Twenty eight percent say they are “not very concerned” (16%) or “not at all concerned” (12%).This bar chart shows the percentage of GitHub users who are very, somewhat, not very, or not at all concerned about the environmental impact (energy and water use and carbon emissions) associated with AI systems (e.g., training and running large language models), or who say they haven't thought about it before. Most GitHub users are concerned about the environmental impact of AI systems. Data: github, Spring 2026. Refer to the data tables in Appendix 1 of the report for all percentages.