<p>Artificial intelligence (AI) has the potential to contribute to environmental protection, as some initiatives have shown; yet, there are some questions about its effectiveness in dealing with environmental challenges and the advancement of sustainable development, as noted by international AI standard-setting instruments. This is because the environment is a complex system, where interactions between natural elements and (mostly local) human dynamics, often driven by inconsistent reasons, can result in unexpected outcomes. Thus, to effectively address environmental issues, it is necessary to consider natural cause–effect relationships as well as human motivations and socio-cultural factors like local practices. The problem is that environmental sustainability efforts based on digital technologies, including AI, are dominated by Eurocentric and technology-based perspectives and conventional sciences, overlooking local knowledge, an approach which may not provide all the answers. To overcome this, it is necessary to rely on other sources of information, in particular Indigenous peoples’ traditional knowledge, which tends to reflect a worldview that favors conservation of natural resources; however, this reliance is not reflected in international law. This paper argues that, by implementing two well-recognized international law principles, public participation and precaution, the consideration of traditional knowledge in the development and building AIs for environmental protection becomes a requirement, as a vital factor in avoiding the materialization of potential harm. This in turn entails the implementation of effective participatory mechanisms for Indigenous peoples and following guidelines that prioritize Indigenous ownership and control over this data, challenging existing colonial approaches embedded in the development of AI technologies.</p>