Advanced

Liability for Copyright Infringement: an investigation of the legal use of trained artificial neural networks in the context of copyright law.

Mutallimzada, Rustam LU (2021) JAEM01 20201
Department of Law
Faculty of Law
Abstract
The current thesis critically investigates whether training and using third-party data in artificial neural networks (ANNs) could constitute a liability for copyright infringement. Furthermore, the present paper sets out to explore whether training and using ANNs could fall within the IP right holders' exclusive rights. In this regard, this thesis will mainly focus on the reproduction parts as the most common infringing restricted act arisen from copying. The thesis also underlines the importance of computational processes that crucially play during the use of trained NNs. Although the infringing acts from NN-generated works are applicable in many other forms of intellectual property law such as design, patent and trademark law, the... (More)
The current thesis critically investigates whether training and using third-party data in artificial neural networks (ANNs) could constitute a liability for copyright infringement. Furthermore, the present paper sets out to explore whether training and using ANNs could fall within the IP right holders' exclusive rights. In this regard, this thesis will mainly focus on the reproduction parts as the most common infringing restricted act arisen from copying. The thesis also underlines the importance of computational processes that crucially play during the use of trained NNs. Although the infringing acts from NN-generated works are applicable in many other forms of intellectual property law such as design, patent and trademark law, the current thesis primarily draws attention from a copyright perspective. At the current stage, the ANN-generated work cannot be commercially exploited no matter whether the work is a copy reproduced from an original copyright work, or whether it is an innovative independent creation of an ANN-system. Additionally, it is essential to examine the requirements that amount to primary and secondary copyright infringements. It is also crucial to analyse whether the infringer's state of mind resulting from secondary copyright infringement can be viewed and applied to ANN-systems. Therefore, the current thesis invites the legislative bodies to reconsider the current traditional copyright laws in relation to ANN-generated works and to set out explicit provisions on the liable party. (Less)
Please use this url to cite or link to this publication:
author
Mutallimzada, Rustam LU
supervisor
organization
course
JAEM01 20201
year
type
H1 - Master's Degree (One Year)
subject
keywords
Artificial Neural Networks Artificial Intelligence Copyright Direct and Indirect Copyright Infringements
language
English
additional info
If you have any questions or concerns about my research paper, please do not hesitate to contact me at: Mutallimzada.rustam@gmail.co
id
9040586
date added to LUP
2021-02-16 17:27:57
date last changed
2021-02-16 17:27:57
@misc{9040586,
  abstract     = {The current thesis critically investigates whether training and using third-party data in artificial neural networks (ANNs) could constitute a liability for copyright infringement. Furthermore, the present paper sets out to explore whether training and using ANNs could fall within the IP right holders' exclusive rights. In this regard, this thesis will mainly focus on the reproduction parts as the most common infringing restricted act arisen from copying. The thesis also underlines the importance of computational processes that crucially play during the use of trained NNs. Although the infringing acts from NN-generated works are applicable in many other forms of intellectual property law such as design, patent and trademark law, the current thesis primarily draws attention from a copyright perspective. At the current stage, the ANN-generated work cannot be commercially exploited no matter whether the work is a copy reproduced from an original copyright work, or whether it is an innovative independent creation of an ANN-system. Additionally, it is essential to examine the requirements that amount to primary and secondary copyright infringements. It is also crucial to analyse whether the infringer's state of mind resulting from secondary copyright infringement can be viewed and applied to ANN-systems. Therefore, the current thesis invites the legislative bodies to reconsider the current traditional copyright laws in relation to ANN-generated works and to set out explicit provisions on the liable party.},
  author       = {Mutallimzada, Rustam},
  keyword      = {Artificial Neural Networks Artificial Intelligence Copyright Direct and Indirect Copyright Infringements},
  language     = {eng},
  note         = {Student Paper},
  title        = {Liability for Copyright Infringement: an investigation of the legal use of trained artificial neural networks in the context of copyright law.},
  year         = {2021},
}