Photo credit: bookriot.com

The Evolving Landscape of Plagiarism in Literature and the Impact of AI

In a recent article in the New York Times, Emily Eakin explored the contentious issue of plagiarism within the literary world, centering her discussion around R. F. Kuang’s 2023 novel Yellowface. The narrative delves deep into themes of cultural appropriation and the privileges associated with race, illustrating a complex relationship between originality and ethics in storytelling. The protagonist, June, appropriates the unpublished work of late author Athena Liu, distorting its essence by introducing racial stereotypes absent from Liu’s original manuscript. Even other characters in the story recognize that June’s narrative does not belong to her, raising important questions about who has the right to tell certain stories. This prompts a broader inquiry into why prominent authors frequently evade consequences for appropriating works from marginalized voices, who often encounter greater obstacles in the publishing landscape.

The landscape of plagiarism has dramatically shifted since the introduction of ChatGPT in late 2022. As awareness and usage of such AI tools grow among students, so does the potential for academic dishonesty. While there is an observable increase in students utilizing ChatGPT for school assignments, it remains uncertain whether this behavior has directly resulted in more instances of outright plagiarism—such as copying without attribution. This ambiguity may suggest that content generated by AI, which often relies on pre-existing works, might obscure the lines of theft. OpenAI admits to training its models on copyrighted works, complicating the ethical usage of these technologies.

Personal experiences with plagiarism inform a deeper emotional connection to this issue. Many authors, including myself, view our writing as an extension of our identities shaped by personal experiences, and the theft of such work feels deeply invasive. Having faced instances where my ideas were duplicated without citation, I witnessed firsthand the inadequacies of current plagiarism detection technologies. Such systems typically access academic databases but often miss material published across diverse platforms, including blogs and personal websites.

Plagiarism can occur unintentionally, often through the improper citation of sources or a lapse in memory regarding where certain ideas originated. In contrast to human oversight, AI lacks genuine understanding or intent, leading to a stricter perception of its actions regarding content generation. The absence of this distinction is concerning, particularly as generative AI bases its outputs on unacknowledged and unlicensed materials.

The ongoing backlash against generative AI is highlighted by a significant ruling in 2023, in which a US court determined that art produced solely by AI lacks copyright protection. While this is a step in the right direction, the ruling’s narrow focus—suggesting that works requiring no human creativity are exempt from copyright—may not encapsulate the complexities of AI usage within creative fields. Additionally, various authors and music creators have pursued class action lawsuits after discovering their works were used to train generative AI models without their consent, advocating for fair compensation for their contributions.

Discussions surrounding whether ChatGPT increases instances of plagiarism in traditional publishing are nuanced. Many systemic issues within the publishing sector predate AI technologies and underscore authors’ legal obligations regarding the originality of their work. Nonfiction writers, for instance, are expected to ensure their books’ accuracy, often employing fact-checkers, research assistants, and sensitivity readers to uphold these standards. Many readers, unfamiliar with these industry practices, mistakenly believe that publishers alone bear the responsibility for ensuring a book’s authenticity and originality.

Alice Nuttall noted in her 2022 analysis that it often requires keenly observant readers to detect instances of plagiarism. Proposals to cross-reference published works against databases for verification on originality could unintentionally promote piracy. Years later, as allegations emerged claiming that OpenAI utilized a repository of pirated materials for training its models, the irony of tools intended to identify plagiarism possibly fostering widespread copyright violations became evident. Furthermore, initiatives to exclusively label human-written books may backfire if these same texts are subsequently used in AI training.

Personally, my apprehension towards generative AI stems from concerns over intellectual property rights and its significant ecological footprint, including high water and electricity usage. The moment is ripe for the introduction of robust regulations focused on technological impact and intellectual property rights rather than diminishing existing protections.

For additional insights into the environmental implications of the publishing industry, further exploration is warranted.

Source
bookriot.com