



Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Community
Ask the community for help and clear up your study doubts
Discover the best universities in your country according to Docsity users
Free resources
Download our free guides on studying techniques, anxiety management strategies, and thesis advice from Docsity tutors
The intricate relationship between copyright law and generative ai. It delves into the ongoing debate surrounding the ownership and protection of ai-generated content, highlighting the challenges and inconsistencies in current legislation. The document examines the impact of training datasets on the legality of ai-created works, the liability concerns surrounding ai actions, and the broader implications for creativity and innovation. It also discusses the recent guidance from the u.s. Copyright office and the need for a more cohesive and unified system to address the evolving landscape of generative ai and copyright. A comprehensive overview of the complex issues at the intersection of ai and intellectual property, offering insights into the future direction of this rapidly changing field.
Typology: Cheat Sheet
1 / 5
This page cannot be seen from the preview
Don't miss anything!
Adrian Serrato 1 ENC Professor Murphy There should not be any copyright protection for AI generated images until they are further altered by the author. There’s no agreement on whether current generation “AI” are sentient or not so the issue of AI owning or having their own copyright protections is not being addressed presently, although it is being considered since it is thought to be a matter of time (Glazer). Copyright law is complex and there are specific variables that can affect the eligibility of an application, including individual decisions by officials on a per application basis. Much of the copyright issues stem from the training sets for generative AI which can sometimes include copyrighted material without permission from the copyright holder(s) (Dubey). There is even debate as to whom is liable when an AI crosses an ethical line; the company that created it, or the user. The issue is complex with many layers, and as time goes on it will have greater and greater impacts on our lives. At its core the purpose of copyright law is to incentivize creativity (Dubey). Copyright law surrounding generative AI deals mainly with copyright of the work made using generative AI and not whether the AI itself should hold copyright protection. The legislation regarding AI authorship is straightforward but, in some cases, unclear. “The Copyright Office declares that it will " register an original work of authorship, provided that the work was created by a human being "…” (Dubey). In other words, one of the minimum requirements to hold a copyright is to be human. AI at the moment are not sentient and so until they are able to achieve consciousness and protest; they are not eligible for copyright protections under current U.S. and much of
European law. International policy in places like India is written less explicitly however, but in practice the outcome is very similar, if not a little inconsistent (Dubey). There are several cases in India that show how inconsistent the system currently is. The Indian copyright office denied an application in which an AI was named sole author, but then accepted an application for another project in which both a human and an AI were listed as the authors after which it then withdrew the decision, the applicant challenged. It is an issue perfectly summarized by, “Put differently, should copyrights be conferred on the maker of the pen or the writer?” (Dubey). Some argue that the “Principal”, or in this case the creators, should hold the copyright and also be liable if the AI does something illegal, for example, copyright infringement (Dubey). Because of how new the technology is, many of the ethical decisions are still being made and as the technology develops so too will the law surrounding it have to evolve to remain effective and relevant. One of the issues complicating the relationship between generative AI and copyright law is the response from people whose jobs it is to protect original copyright material; reacting as the material they oversee, or created, is used in massive training sets to train an AI without consent or credit (Dubey). The datasets used to train generative AI are often very large, consisting of many thousands of images and copyrighted material. There is an interesting correlation that supporters of generative AI copyright point to in that artists that made collages would often use copyrighted images or logos. In the making of these collages, they have changed the images enough so as to make them qualify as their own imagination. They suggest that the way in which AI learns and creates is comparable to the collages made by artists. Therefore, deserving copyright protection in their eyes (Day). Some of the artists and copyright holders who suspect their work has been used in training sets for these AI without consent or credit have sued the AI
generative AI are made, older laws are used as reference. Policy in the area of AI is in development as things evolve, with more advancements will come more changes. With all of this in mind, until a more cohesive and unified system is in place to decide on copyright protections for generative AI, the work it creates, and how it relates to the human that gave the input parameters, images that are entirely AI generated without any post-generation edits should be public domain for the time being. Until a clearer understanding of the way generative AI “learns” and “creates” and its relation to how we as humans learn and create, it’s difficult to say with any conviction that use of copyright material in training sets is copyright infringement because of the apparent similarities between machine learning (AI) and human learning. The question remains, how is the way in which generative AI “learns” and “creates” different from our way of experiencing the world, taking input sensory data and separating parts of the experience, sights, smells, sounds, etc. When we create, aren’t we to some degree pulling from a library of experiences and data we have collected through exploration or experience. There is no clear consensus, but as technology develops the discussion continues. While some of these things seem irrelevant, and are right now, the impacts will be felt in many different industries as AI gets integrated into everything and continues to grow.
Works Cited "Rise of the machines: AI and Copyright." Bar & Bench , 31 Mar. 2023, p. NA. Gale General OneFile , link.gale.com/apps/doc/A743810226/ITOF?u=lincclin_bwcc&sid=bookmark- ITOF&xid=bd2d8af2. Accessed 24 July 2023. Naqvi, Zack. "ARTIFICIAL INTELLIGENCE, COPYRIGHT, AND COPYRIGHT INFRINGEMENT." Marquette Intellectual Property Law Review , vol. 24, no. 1, Wntr 2020, pp. 15+. Gale General OneFile , link.gale.com/apps/doc/A640360965/ITOF? u=lincclin_bwcc&sid=bookmark-ITOF&xid=c149b8ab. Accessed 24 July 2023. Glazer, Sarah. "The Future of Artificial Intelligence." CQ Researcher, 25 Nov. 2022, pp. 1-31, library.cqpress.com/cqresearcher/cqresrre2022112500. Day, Kathleen. "Artificial Intelligence and Intellectual Property." CQ Researcher, 21 Apr. 2023, pp. 1-19, library.cqpress.com/cqresearcher/cqresrre2023042102. "U.S. Copyright Office Provides Guidance On Protecting AI-Created Works." Mondaq Business Briefing, 29 Mar. 2023, p. NA. Gale General OneFile, link.gale.com/apps/doc/A743411085/ITOF?u=lincclin_bwcc&sid=bookmark- ITOF&xid=97b5248d. Accessed 30 July 2023.