This article highlights the environmental impacts associated with information and communication technologies (ICT) used for data storage and processing, emphasizing the significant emissions generated during the lifecycle of electronic devices. It addresses the challenges of assessing these environmental impacts using methodologies like the GHG Protocol and Life Cycle Assessment (LCA). It also explores opportunities for mitigating these impacts through better data governance, techniques for reducing digital waste, and sustainability initiatives such as the Arch’Eco project, who aims to assess environmental impacts for the entire lifecycle of data and identify best practices for data management in an environmentally friendly manner.

The increasing volume of generated data and archives raises pertinent questions regarding the effectiveness of traditional archival appraisal methods, which largely depend on human expertise. As automation and artificial intelligence (AI) become prevalent in various sectors, the field of archiving stands on the brink of significant transformation.This paper explores the integration of AI within archival appraisal processes, framed within the context of the Maturity Assessment for Appraisal (MAA) project (2023-2025). The MAA seeks to evaluate the defensibility, stability, and appropriateness of current appraisal practices while assessing the readiness of records for automated appraisal. Employing exploratory qualitative research, the study outlines a systematic approach that includes a literature review, testing of a maturity model, and consultations with archival professionals. The MAA encompasses six key dimensions: principles, vision/strategic framework, compliance, methodology, tools, and criteria, providing a structured framework for assessing the maturity of appraisal practices in the AI age. Preliminary results highlight the potential benefits of AI in enhancing appraisal efficiency and effectiveness, paving the way for more informed and defensible archival decisions. Some applied use cases are already started and primarily derived results will be highlighted