@article{ART002784242},
author={Yoon, Kyung-Sub},
title={A Review of Design Defects and Product Liability of AI Networks},
journal={Legal Theory & Practice Review},
issn={2288-1840},
year={2021},
volume={9},
number={4},
pages={189-232}
TY - JOUR
AU - Yoon, Kyung-Sub
TI - A Review of Design Defects and Product Liability of AI Networks
JO - Legal Theory & Practice Review
PY - 2021
VL - 9
IS - 4
PB - The Korea Society for Legal Theory and Practice Inc.
SP - 189
EP - 232
SN - 2288-1840
AB - This study mainly listened to the issues of product liability law that are associated with the AI development guideline as an example. As it is an example, it does not cover everything, and there are many parts that only raise issues. In the first place, since the subject of this paper, the AI network, is still in its cradle, it is difficult to predict everything accurately, and there are limitations in the abilities of researchers, so please understand that there are many shortcomings. If I have to mention two important points last, one would be that the design flaws of products using AI are exposed to ex post evaluation by comparison with RAD.
Manufacturers and others who are considering using AI in their products should fully understand that AI has drawbacks such as “uncontrollability” and “opaqueness”. It should then be considered as far as possible to adopt possible design measures to minimize the shortcomings. Otherwise, in the event of an accident, there is a high risk of design defects being recognized because RAD (a suitable alternative design plan) was not employed. The second important point is the principle of malfunction.
Although the plaintiff cannot prove by direct evidence that the specific defect in the product using AI and that defect was the cause of the damage, if the plaintiff uses the malfunction principle, it can be ratified by the court through indirect circumstantial evidence. Although AI lacks controllability and transparency, even if a defect or causal relationship cannot be specified for the plaintiff, there may be cases in which manufacturers, etc. cannot escape product liability if the malfunction law is applied. Therefore, manufacturers, etc., may have to request that defects such as uncontrollability and opacity be cured or improved as much as possible when receiving AI supply.
KW - artificial intelligence;design flaws;product liability;malfunction principle;controllability;transparency principle;no-fault.
DO -
UR -
ER -
Yoon, Kyung-Sub. (2021). A Review of Design Defects and Product Liability of AI Networks. Legal Theory & Practice Review, 9(4), 189-232.
Yoon, Kyung-Sub. 2021, "A Review of Design Defects and Product Liability of AI Networks", Legal Theory & Practice Review, vol.9, no.4 pp.189-232.
Yoon, Kyung-Sub "A Review of Design Defects and Product Liability of AI Networks" Legal Theory & Practice Review 9.4 pp.189-232 (2021) : 189.
Yoon, Kyung-Sub. A Review of Design Defects and Product Liability of AI Networks. 2021; 9(4), 189-232.
Yoon, Kyung-Sub. "A Review of Design Defects and Product Liability of AI Networks" Legal Theory & Practice Review 9, no.4 (2021) : 189-232.
Yoon, Kyung-Sub. A Review of Design Defects and Product Liability of AI Networks. Legal Theory & Practice Review, 9(4), 189-232.
Yoon, Kyung-Sub. A Review of Design Defects and Product Liability of AI Networks. Legal Theory & Practice Review. 2021; 9(4) 189-232.
Yoon, Kyung-Sub. A Review of Design Defects and Product Liability of AI Networks. 2021; 9(4), 189-232.
Yoon, Kyung-Sub. "A Review of Design Defects and Product Liability of AI Networks" Legal Theory & Practice Review 9, no.4 (2021) : 189-232.