Publication
Sustainable Accessible Artificial Intelligence (Sustainable zug.KI)
Daniel Sonntag; Hannes Kath; Christoph Albert Johns; Thiago Gouvea
German Research Center for AI, DFKI Technical Report, 2026.
Abstract
The 'Sustainable zug.KI' project, funded by the Lower Saxony Ministry of Science and Culture (2025-2028), aims to develop robust methods to quantify the diverse effects of explainable artificial intelligence (XAI) on sustainability across ecological, economic, and societal dimensions. By incorporating XAI techniques into machine learning, knowledge representation, and intelligent user interfaces, the project strives to facilitate the creation of more sustainable systems. An essential goal is to ensure that the complete AI lifecycle - ranging from development and deployment to usage and reuse - is scrutinised from a sustainability perspective. In response to the escalating demand for AI systems that maintain both transparency and trustworthiness, the project is primarily concentrated on resource-limited areas, such as ecological monitoring. By leveraging interactive machine learning (IML), Sustainable zug.KI intends to generate tools that enable domain specialists, including ecologists, to actively engage in the machine learning process. Moreover, the project will devise efficient methods for transferring AI knowledge, thereby fostering interdisciplinary cooperation and enhancing the practical implementation of AI in sustainability initiatives. Sustainable zug.KI aspires to further both foundational research and real-world applications, contributing to global efforts to fulfil sustainability goals through responsible and explainable AI.