Federated learning, ethics, and the double black box problem in medical AI
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911577647611904 |
|---|---|
| author | Hatherley, Joshua Søgaard, Anders Ballantyne, Angela Pauwels, Ruben |
| author_facet | Hatherley, Joshua Søgaard, Anders Ballantyne, Angela Pauwels, Ruben |
| contents | Federated learning (FL) is a machine learning approach that allows multiple devices or institutions to collaboratively train a model without sharing their local data with a third-party. FL is considered a promising way to address patient privacy concerns in medical artificial intelligence. The ethical risks of medical FL systems themselves, however, have thus far been underexamined. This paper aims to address this gap. We argue that medical FL presents a new variety of opacity -- federation opacity -- that, in turn, generates a distinctive double black box problem in healthcare AI. We highlight several instances in which the anticipated benefits of medical FL may be exaggerated, and conclude by highlighting key challenges that must be overcome to make FL ethically feasible in medicine. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_20656 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Federated learning, ethics, and the double black box problem in medical AI Hatherley, Joshua Søgaard, Anders Ballantyne, Angela Pauwels, Ruben Machine Learning Artificial Intelligence Computers and Society Human-Computer Interaction Federated learning (FL) is a machine learning approach that allows multiple devices or institutions to collaboratively train a model without sharing their local data with a third-party. FL is considered a promising way to address patient privacy concerns in medical artificial intelligence. The ethical risks of medical FL systems themselves, however, have thus far been underexamined. This paper aims to address this gap. We argue that medical FL presents a new variety of opacity -- federation opacity -- that, in turn, generates a distinctive double black box problem in healthcare AI. We highlight several instances in which the anticipated benefits of medical FL may be exaggerated, and conclude by highlighting key challenges that must be overcome to make FL ethically feasible in medicine. |
| title | Federated learning, ethics, and the double black box problem in medical AI |
| topic | Machine Learning Artificial Intelligence Computers and Society Human-Computer Interaction |
| url | https://arxiv.org/abs/2504.20656 |