In Part 1 of this series of articles: “The Search for a Wider Understanding Regarding Authorship and Fair Use,” I looked at the recent Bartz v Anthropic and Suno decisions.
In this article, I dive a bit deeper into the AI/authorship issue and explore some of the unique challenges we face.
An Introduction
My underlying argument is that the whole notion of authorship has to be reassessed. However, before we do that, we need to acknowledge the enormity of the process of change that humanity is going through at the moment. It is both pervasive and unprecedented. Take for example the way we think and communicate. In a recent article in The Conversation,[i] Antonio Cerella discusses how AI is damaging our ability to use language effectively and to think creatively. Children have great imaginations. As Cerella writes, they live in a world full of sensations, but they seldom have the words to describe those sensations. That is where language comes into play. As we grow older, we learn to express ourselves and make sense of the world around us. We do this through use, and in rare cases, our mastery of language, for example, with music, the lyrics of a song.
Cerella explains that an amazing capacity for abstraction liberates us from the immediacy of experience. However, he also cautions that, at the same time, it is a fragile relationship that “can decay when language is dictated rather than discovered”, explaining that “The result is a culture of immediacy, dominated by emotion without understanding, expression without reflection.” Once analysed, what he is in fact describing is the brave new world that we are all getting dragged into, whether we like it or not. It is a fast-moving tech-driven world where AI provides us with a tool to avoid time-consuming tasks and provides us with quick and easy solutions. It also offers us the opportunity for a range of human and machine collaboration, unimaginable[ii] even a decade ago. That is, the ability to collaborate and co-author all manner of creative works with non-human intelligence. This is the issue I will look at in a couple of posts.
In January 2025, the U.S. Copyright Office determined[iii]that AI-generated content can attract copyright protection, if there is a substantial degree of human intervention in the creative process. Some commentators see this as a turning point in AI-augmented music production, opening doors to new possibilities for creators, while maintaining the importance of human authorship. Others warn that the real problem area is identifying who owns the copyright when a work such as a song is composed by AI. The difficulty that arises: how do we identify to what extent human input and participation is essential for copyright consideration and where do we draw the line drawn between AI assistance and full human authorship?
These are hard questions to answer. The problem, on the other hand, is easy to identify. It is that mind-blowingly large quantities of copyrighted works are being created daily. For example, it is estimated that 100,000 new songs are loaded onto streaming services each day[iv]. Undoubtedly, rapidly increasing number of these works are created through shared AI and human involvement. It is therefore an issue we must address and deal with now, not at some stage in the future.
Looking at the music industry as an example, the immediate problem is that today sufficiently large quantities of music are being composed, produced, and arranged by AI, without any human intervention. The estimates vary but even if we assume it is 10% of the 100,000 new tracks that emerge each day, that is 10,000 tracks a day. The streaming platform Deezer estimates[v] that as at September 2025, 30,000 AI tracks are uploaded daily, representing 28% of all uploads. At present, these works do not attract copyright. It means they exist in some sort of legal vacuum, existing in the real world, and in real terms, insofar as they can be accessed on streaming platforms like Spotify. However, because they are not deemed to be copyright works, they can be reproduced, remixed, in whole or part and distributed freely without infringing anyone’s rights.
The Conundrum of Cognitive Debt
In my first post I looked at how AI is testing the boundaries of established copyright doctrine. We have been debating for some time now how the law needs to be reshaped to address AI-generated content. At a recent conference “Beyond Boundaries: Code to Create”[vi] Professor Ryan Abbott, a world leading authority on the topic discussed the challenges of determining and assigning authorship in the age of generative AI. Abbott referred to the United Kingdom’s provision for computer-generated works, the same provision New Zealand adopted, which recognises that computers and humans can and often do collaborate to create original works, and that the creative process may be a shared one. However, he describes efforts to achieve international co-ordination and regulation, as “all a bit confused right now”. With respect, that might be a bit of an understatement. A more apt description would be that, with a few exceptions, we are not just confused, but we have our heads firmly in the sand.
At the same conference, the panelists addressed the problem of the sheer volume of AI-generated content that is flooding the internet. Lin Wu, the head of regional legal and ecommerce legal teams at TikTok, explained how the platform manages AI-generated content. It requires creators to label realistic AI-generated content, whether in the form of an image, video or audio, that has been entirely or significantly altered or created by AI. TikTok also has its own automated detection systems to tag such content. This assists users to distinguish between human and AI created content. It is a start, but we need to do a lot more.
Before looking at things in detail we need to step back and look at the way AI is impacting how we act and think. Recent neuroscientific research reveals a fundamental shift in how AI changes the creative process. A groundbreaking MIT study tracked students’ neural and behavioural responses while using ChatGPT for essay writing tasks.[vii] The researchers discovered that AI assistance appeared to dampen certain brain activities, leading to what they termed “cognitive debt”. That is, a concerning reduction in the mental effort required for complex tasks.
Similarly, researchers at the University of Pennsylvania’s Wharton School compared cognitive outcomes between two groups: those who used large language models like ChatGPT for research, versus those who relied on traditional Google searches. The results were striking. Participants who used AI chatbots developed what researchers diplomatically called “shallower knowledge” compared to their counterparts who navigated traditional search results themselves.
The root of this cognitive shift lies in how AI fundamentally changes our relationship with information. Traditional search engines presented us with links that required active evaluation, synthesis, and critical thinking. We had to read multiple sources, compare perspectives, and draw our own conclusions. Some conclusions were better than others, but importantly they were our own. AI tools, by contrast, serve up pre-digested summaries and analyses. While this increases efficiency, it reduces our active engagement with source material. We become passive consumers rather than active processors of information. The consequence extends beyond mere convenience. We are losing the mental exercise that builds critical thinking skills and deep understanding.
The MIT study “Your Brain on ChatGPT” demonstrates that AI assistance reduces neural activity in brain regions associated with executive control, language processing, and idea generation. Key findings show that students using AI showed significantly lower brain connectivity compared to those writing independently, memory retention of AI-generated content was markedly reduced, suggesting a shift from internalised to externalised cognitive processes, and the timing of AI intervention matters: early AI use diminishes cognitive engagement, while later incorporation maintains stronger neural activity. Crucially, this is not merely about cognitive decline but has potentially important ramifications for the whole question of human authorship.
The discussion of the concept of “cognitive debt” and the use of empirical data to properly understand it has profound implications for copyright law’s assumptions about human creativity and the cognitive processes underlying original expression. As the authors of an article “How ChatGPT May Be Impacting Your Brain” observe, “Letting AI write or think for you may feel efficient but could undermine long-term critical thinking and creativity”.[viii] That is a concern in terms of creativity. It is likely to impact how concepts of authorship are shaped in the future, because if we stop rewarding creativity, we undermine the whole premise of protecting creators and the spark of genius that they are often able to exhibit.
===============================================================
[i] How generative AI could change how we think and speak, Antonio Cerella, The Conversation,
25 October 2025, https://share.google/aUMjWlrw8OTna3EEz
[ii] Other than in an Isaac Asimov or Arthur C Clarke science-fiction novel
[iii] U.S. Copyright Office, Copyright and Artificial Intelligence, Part 2: Copyrightability, January 2025
[iv] https://www.gearnews.com/spotify-streaming-report-2024-tech/
[v] https://newsroom-deezer.com/2025/09/28-fully-ai-generated-music/
[vi] IP Week @ SG: Experts debate AI authorship, copyright and creative ownership
01 September 2025, https://asiaiplaw.com/section/news-analysis/ip-week-at-sg-experts-debate-ai-authorship-copyright-and-creative-ownership
[vii] Kosmyna, N., Hauptmann, E., Yuan, Y., Situ, J., Liao, X.-H., Beresnitzky, A., Braunstein, I., & Maes, P. (2025). Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task. arXiv:2506.08872
[viii] Marlynn Wei, “How ChatGPT May Be Impacting Your Brain,” Psychology Today, June 19, 2025, https://www.psychologytoday.com/nz/blog/urban-survival/202506/how-chatgpt-may-be-impacting-your-brain
