Recent proposals to use artificial intelligence (AI) in end-of-life decision-making for incapacitated patients without advance directives have prompted critical reflection by the Ethics Committee of the Italian Society of Anesthesia, Analgesia, Resuscitation and Intensive Care (SIAARTI). This position paper analyzes both general ethical concerns surrounding AI in clinical practice and specific issues raised by the hypothesis of using AI to reconstruct patients’ presumed wishes through recorded clinical conversations or analysis of digital footprints. At a general level, major challenges include lack of explainability (the so-called black box problem), risks of bias linked to non-representative training data, environmental sustainability, and the potential erosion of the clinician-patient relationship. In end-of-life contexts, these concerns are amplified. Systematic recording and retrospective analysis of sensitive conversations raise serious questions regarding privacy, data security, informed consent, and the authenticity of communication. Moreover, algorithmic interpretation may fail to capture the complexity of non-verbal communication and the inherently interpretative nature of moral reasoning. The construction of a “social portrait” from digital traces risks oversimplifying personal identity and generating conflicts with family narratives at a particularly vulnerable moment. The Committee further highlights the risk of delegating ethically weighty relational tasks to technological systems, thereby reinforcing a procedural model of medicine and weakening shared decision-making. For these reasons, the proposed use of AI in this domain is considered ethically problematic in its current form. Any future application would require robust governance, transparency, and accountability, ensuring that AI supports rather than undermines authentic care relationships.

The Ethics Committee of the Italian Society of Anesthesia, Analgesia, Resuscitation and Intensive Care (SIAARTI) — Artificial intelligence in end-of-life decision-making processes: ethical reflections

Furlan, Enrico
;
Piccinni, Mariassunta;
2026

Abstract

Recent proposals to use artificial intelligence (AI) in end-of-life decision-making for incapacitated patients without advance directives have prompted critical reflection by the Ethics Committee of the Italian Society of Anesthesia, Analgesia, Resuscitation and Intensive Care (SIAARTI). This position paper analyzes both general ethical concerns surrounding AI in clinical practice and specific issues raised by the hypothesis of using AI to reconstruct patients’ presumed wishes through recorded clinical conversations or analysis of digital footprints. At a general level, major challenges include lack of explainability (the so-called black box problem), risks of bias linked to non-representative training data, environmental sustainability, and the potential erosion of the clinician-patient relationship. In end-of-life contexts, these concerns are amplified. Systematic recording and retrospective analysis of sensitive conversations raise serious questions regarding privacy, data security, informed consent, and the authenticity of communication. Moreover, algorithmic interpretation may fail to capture the complexity of non-verbal communication and the inherently interpretative nature of moral reasoning. The construction of a “social portrait” from digital traces risks oversimplifying personal identity and generating conflicts with family narratives at a particularly vulnerable moment. The Committee further highlights the risk of delegating ethically weighty relational tasks to technological systems, thereby reinforcing a procedural model of medicine and weakening shared decision-making. For these reasons, the proposed use of AI in this domain is considered ethically problematic in its current form. Any future application would require robust governance, transparency, and accountability, ensuring that AI supports rather than undermines authentic care relationships.
File in questo prodotto:
File Dimensione Formato  
unpaywall-bitstream-797706034.pdf

accesso aperto

Tipologia: Published (Publisher's Version of Record)
Licenza: Creative commons
Dimensione 814 kB
Formato Adobe PDF
814 kB Adobe PDF Visualizza/Apri
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11577/3597086
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? 0
  • OpenAlex ND
social impact