TY - JOUR
T1 - The proposed EU Directives for AI liability leave worrying gaps likely to impact medical AI
AU - Duffourc, Mindy Nunez
AU - Gerke, Sara
N1 - This work was funded by the European Union (Grant Agreement no. 101057321). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the Health and Digital Executive Agency. Neither the European Union nor the granting authority can be held responsible for them. S.G. also reports grants from the European Union (Grant Agreement no. 101057099), the National Institute of Biomedical Imaging and Bioengineering (NIBIB) of the National Institutes of Health (Grant Agreement no. 3R01EB027650-03S1), and the Rock Ethics Institute at Penn State University.
This work was funded by the European Union (Grant Agreement no. 101057321). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the Health and Digital Executive Agency. Neither the European Union nor the granting authority can be held responsible for them. S.G. also reports grants from the European Union (Grant Agreement no. 101057099), the National Institute of Biomedical Imaging and Bioengineering (NIBIB) of the National Institutes of Health (Grant Agreement no. 3R01EB027650-03S1), and the Rock Ethics Institute at Penn State University.
PY - 2023/12
Y1 - 2023/12
N2 - Two newly proposed Directives impact liability for artificial intelligence in the EU: a Product Liability Directive (PLD) and an AI Liability Directive (AILD). While these proposed Directives provide some uniform liability rules for AI-caused harm, they fail to fully accomplish the EU’s goal of providing clarity and uniformity for liability for injuries caused by AI-driven goods and services. Instead, the Directives leave potential liability gaps for injuries caused by some black-box medical AI systems, which use opaque and complex reasoning to provide medical decisions and/or recommendations. Patients may not be able to successfully sue manufacturers or healthcare providers for some injuries caused by these black-box medical AI systems under either EU Member States’ strict or fault-based liability laws. Since the proposed Directives fail to address these potential liability gaps, manufacturers and healthcare providers may have difficulty predicting liability risks associated with creating and/or using some potentially beneficial black-box medical AI systems.
AB - Two newly proposed Directives impact liability for artificial intelligence in the EU: a Product Liability Directive (PLD) and an AI Liability Directive (AILD). While these proposed Directives provide some uniform liability rules for AI-caused harm, they fail to fully accomplish the EU’s goal of providing clarity and uniformity for liability for injuries caused by AI-driven goods and services. Instead, the Directives leave potential liability gaps for injuries caused by some black-box medical AI systems, which use opaque and complex reasoning to provide medical decisions and/or recommendations. Patients may not be able to successfully sue manufacturers or healthcare providers for some injuries caused by these black-box medical AI systems under either EU Member States’ strict or fault-based liability laws. Since the proposed Directives fail to address these potential liability gaps, manufacturers and healthcare providers may have difficulty predicting liability risks associated with creating and/or using some potentially beneficial black-box medical AI systems.
UR - http://www.scopus.com/inward/record.url?scp=85156129320&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85156129320&partnerID=8YFLogxK
U2 - 10.1038/s41746-023-00823-w
DO - 10.1038/s41746-023-00823-w
M3 - Article
C2 - 37100860
AN - SCOPUS:85156129320
SN - 2398-6352
VL - 6
JO - npj Digital Medicine
JF - npj Digital Medicine
IS - 1
M1 - 77
ER -