Plagiarism in the World of Generative AI

Image by kenshinstock on Freepik.

By Rachel Guise, PPS Editor

Generative artificial intelligence (AI) has become incredibly popular in 2023 through TikTok trends that turn selfies into portraits from other centuries and different fantasies to Snapchat’s My AI chatbot. This new technology has greatly affected all aspects of the creative industry. Let’s discuss how generative AI affects the publishing industry.

First, a definition of generative AI: According to IBM Research, “Generative AI refers to deep-learning models that can generate high-quality text, images, and other content based on the data they were trained on.” So it begs the question, if content is being created from the information provided, wouldn’t this classify as plagiarism?

The short answer is yes. According to an article by The Guardian, plagiarism is harder to detect when written by AI because interfaces such as ChatGPT are writing at higher levels. The article describes a situation in which an academic paper, written by ChatGPT, went undetected during a peer review as the reviewers assumed it was written by the scholars who posed the original experiment.

Furthermore, the Authors Guild proposed a class action lawsuit against one of the top generative AI companies, Open AI. Authors were curious about how well ChatGPT would be able to describe one of their characters based off a snippet of their book. They were surprised—and displeased—to find out the chatbot was able to write very detailed information, more so than they thought would be possible from the information they provided. Upon further research, they discovered that their existing books had been poured into the platform as content to be used for generative AI writing without their knowledge, consent, or approval.

So as a user of generative AI, how can you sift through the generated content to determine if it was essentially copy and pasted from another source?

How to Identify

Copyleaks is one of the top resources to identify plagiarism with both an AI Content Detector and Plagiarism Detector available for use. According to their website, the AI Content Detector is “[t]he only enterprise solution that keeps you informed about what content was written by a human or generated by AI, including GPT-4 and Bard.” Their Plagiarism Detector promises to “identify potential plagiarism, detect AI-generated content, and more.”

As an author, there are searchable databases available to see if your content has been used for generative AI, as noted by Owlcation.

How to Prevent

The most obvious solution would be to avoid using any type of generative AI software when writing. But if you need a little help, it would be beneficial to make use of tools such as Copyleaks to help identify any potentially plagiarized material you use from a generative AI platform.

If you happen to be affected by AI and notice your work was illegally being used, one solution is to send a letter to AI companies alerting them your work is being used without your consent, which breaks copyright laws. This will aid in taking your work out of their databases.