Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
European Product Liability for AI-based Clinical Decision Support Systems
2
Zitationen
1
Autoren
2024
Jahr
Abstract
Abstract Clinical decision support systems (CDSSs) are computer systems that support medical professionals in the clinical decision-making process. If a decision that involved the use of a CDSS has detrimental effects, the patient might seek compensation for the damage incurred. Here, the involvement of a CDSS, particularly if it is AI-based, could give rise to liability questions. On the one hand, the liability of the medical professional, which often depends on medical standards, raises questions on the type of behaviour on which these standards will require, particularly for AI-based CDSSs. On the other hand, questions arise on the liability of the manufacturer of the system. Product liability regimes such as those in the US and the EU do not require the claimant to prove fault, but instead require a product defect. For these regimes, AI-based CDSSs challenge the notions of defect and causation. This chapter focuses on EU product liability law, analysing whether and how it applies to AI-CDSSs. The three main questions discussed are (i) whether AI-CDSSs could fall within the scope of this regime, (ii) how the term ‘defect’ should be construed with respect to AI-CDSSs, arguing for a novel interpretation of AI manufacturing defects, and (iii) whether and how a causal link between defect and damage could be established.
Ähnliche Arbeiten
Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI
2019 · 8.402 Zit.
Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead
2019 · 8.270 Zit.
High-performance medicine: the convergence of human and artificial intelligence
2018 · 7.702 Zit.
Proceedings of the 19th International Joint Conference on Artificial Intelligence
2005 · 5.781 Zit.
Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI)
2018 · 5.507 Zit.