GenAI content raises questions about ownership as copyrighted works have usually been associated with humans. This collaboration with human and machine blurs the lines of ownership and therefore presents a conundrum for copyright law.
The article discusses debates which centre on the point that AI lacks human authorship and thus should be exempt from protection whilst have argued that since it does have human authorship at the origin including data sets and refining algorithms it should be copyrighted.
Bahodirugli says some suggest giving AI some form of person hood, others suggest expanding the ownership of copyright to include AI systems. The key discussion centres on finding the balance between human creativity and AI innovation and technology and the importance of navigating the complexities in order to develop a legal framework.
One of the challenges is attributing ownership to AI as an independent author since copyright provides ownership to authors to allow them to disseminate their work. This is complicated in the case of AI as it is a collaboration between humans and AI. There is thus the need to re-evaluate the ideas around copyright itself.
He discusses the concept of ‘Fair Use’ where copyrighted works can be used for things such as analysis, parody or commentary. This is problematic when it comes to AI as AI can generate content using copyrighted content and raises questions as to at what point can this be fair use or copyright infringement.
Another concept causing problems is called ‘Sweat of the brow’, which deals with the fact that effort is put into the creation of works, but AI once again blurs this line as the questions around using GenAI content and human effort and labour come into question.
There is the discussion around bias caused by human input of data which perpetuate society’s bias. The article suggests this requires scrutiny of the training data which is a good idea but society is inherently biased and developers will have to be intentional about reducing bias in their tools which is easier said than done especially if there are financial implications.
The article calls for dialogue between policy makers, legal experts, AI developers, ethicists and other stakeholders to collaborate to address the AI/human copyright issues in order to protect human ownership whilst still giving space to innovation in AI technology and ensuring informed decision making when it comes to developing regulation. Bahodirugli mentions the importance of human-AI collaboration for better efficiency and innovative creation and calling for interdisciplinary research and collaboration with all stakeholders developing regulatory frameworks whilst protecting the rights of creatives and encouraging innovation.
The article refers to the ‘Black Box’ of AI generated content and call for transparency to ensure people are aware when they interact with GenAI content. The ‘Black Box’ they say refers to the bias, unintended harm and lack of explainability. The issue for me is generalised words like these are irritating and cause fear and trepidation for people without definitive proof, words such as ‘unintended harm’. These words cannot be used unless they are clearly defined and explained—in what context, how, when and so on.
Bahodirugli discusses issues of privacy as AI analyses and uses large amounts of private data. They suggest policy makers and develops should work together to create guidelines around the use of personal data in AI.
He suggests that awareness campaigns and educational initiatives should be implemented to encourage public acceptance and understanding of AI tools and suggests that people should be educated about the ‘potential loss of human creativity’ due to AI. Another generic comment which makes no sense and is untrue. Human beings have creativity naturally It cannot be taken away by AI or anyone else as it is inherent to people.
Conclusion:
An interesting and educative article to read in understanding the issues around copyright in relation to AI and human generated content.
The need for education and for ensuring all stakeholders come to the table have also been called for by various stakeholders such as the UK’s House of Lords (Tobin, 2022) and the Council of Europe, (Holmes, 2023). A regulatory framework is slow in coming and the longer government and policy holders take to develop a framework for developers, the more society is unsure about how to interact with AI.
The regulatory framework should not stifle AI’s development or innovative capacity as the use of the internet is a precursor as to how people interact within digital spaces and use digital tools, but in terms of ownership and copyright, there is a need to protect originators of content by developers and policy makers, and, this conversation is best had sooner rather than later to provide a sense of safety for creatives about their work.
Review by Nana Ofori-Atta Oguntola
Unraveling the Copyright Conundrum: ExploringAI-Generated Content and its Implications for Intellectual Property Rights AbdikhakimovIslombek Bahodirugli, science Zone https://www.science-zone.org/index.php/conference/article/view/42/37
Comentarios