On the Interplay of Legal Requirements, Quality Aspects and Ethical Risks when using Machine Learning in German Official Statistics
Articles and reports: 11-522-X202500100014Description: Artificial intelligence (AI) with its subfield machine learning (ML) has found its way into administration in general and also into official statistics in Germany in particular. This paper highlights the ethical issues that may arise when using AI/ML in official statistics and examines whether a separate ethical framework is needed to deal with these issues appropriately, as is proposed by institutions of other countries and intergovernmental institutions related to official statistics. The results of the study are presented to show that the implementation of the requirements of the existing and mostly non-AI/ML-specific frames of reference such as law and quality is already sufficient to adequately address the ethical issues based on risk scenarios.
Issue Number: 2025001Author(s): Dumpert, FlorianMain Product:Statistics Canada International Symposium Series: Proceedings
