3 Strange Ways AI is Improving Plagiarism

When it comes to plagiarism and authorship, there is little doubt that the rise of generative AI has been the most significant turning point since the advent of the internet itself.

For the first time, it is trivial to generate large amounts of text (or other content) that, while imperfect, is often passable for human work. This has sent shockwaves through creatives and academics alike.

On the whole, there’s little doubt that generative AI has significantly increased the prevalence of plagiarism both online and in the classroom. The amount of content being presented as human-created, when in fact it’s AI-generated, is rising.

Then there are the issues with AI itself. Just yesterday, I discussed how ChatGPT rehashed an article I’d written, even though it was supposedly barred from indexing my site. AI is deeply problematic both in how it is trained and how it is being used.

But no change comes without some upsides. Although it may be challenging to find, there are small ways AI is improving authorship. As such, I feel it’s important to acknowledge at least some of the less obvious but more positive impacts the technology is having.

Even if they don’t outweigh the issues raised, it is still important to be aware of these changes.

1: Less Plagiarism of Human-Created Work

When I first became interested in plagiarism during the early 2000s, it was because of widespread plagiarism of my work. This included poetry, short stories, essays and other content I had published online.

This wasn’t just a problem for creatives. Businesses had their sites copied by upstart competitors, and academics had their work copied without attribution by later researchers. Up until very recently, if you wanted to plagiarize, it had to come at the expense or with the permission of a human author.

For many creators, this is not an improvement. It simply converts occasional direct plagiarism into regular indirect plagiarism, with AI systems serving as the middleperson in the exchange. But concerns about search engine optimization (SEO), in particular duplicative content, and other direct impacts of the plagiarism are less common.

However, that is cold comfort. Given the widespread reduction in traffic from search engines because of AI, any SEO benefit is likely far overshadowed by the other impacts.

Still, there is far less direct plagiarism of humans. While it does still happen, much of the plagiarism has pivoted to AI.

2: The Decline of Essay Mills

Essay mills and other contract cheating services have been in sharp decline since late 2022. Chegg, one of the largest “homework helper” websites has been teetering on collapse and has even sued Google over AI summaries.

Though paid contract cheating is still exists, it’s also struggling. How do you motivate a student to pay for a service that is slower and likely inferior to what an AI can provide for free?

Though some may be motivated by an attempt to avoid AI detectors, essay mills have a history of not doing the work they are paid for. How do you trust the human author didn’t just use AI themselves?

If a student is determined to cheat, AI is the way to go. For academia, this feels like swapping out one enemy for another that’s even more difficult to detect. But essay mills were exploitative in ways in a myriad of ways that AI tools are not.

Generative AI creates a slew of problems for teachers, but at least it isn’t deliberately trying to sabotage education through exploitative marketing and spammy behavior. Furthermore, essay mills have no potential for a positive relationship with education because their entire purpose is to short circuit assessment. With AI, at least there are some positive uses for teachers and students alike.

This at least gives some hope that the problem can potentially be addressed and some form of symbiosis can be reached.

3: A Real Conversation About Authorship

AI has forced a conversation that has been lingering in the background for decades: what does it mean to be the author of a work?

This question has taken many, many forms over the years. How much should a tutor correct a student paper? When is it acceptable to use a ghostwriter? How much editing is too much and fundamentally changes a work? When does a person qualify to be an author of a research paper?

Even copyright law has its own version of this queastion, asking when a copy of the orignal, a derivative work or an entirely new creation?

This conversation has always been a part of creation. As long as there been coauthors, editors, sources of inspiration and other types of collaborators. However, technology has been pushing it further and further. What began with automated spell checking and grammar checking has ballooned into major rewrites and restructuring.

Nearly all works, including this article, are collaborations between humans and technology. We’ve largely ducked the question about when the line of authorship is crossed. However, with generative AI, that question is at the forefront and is not going away.

Bottom Line

For cynics and critics of AI, these positives do not come anywhere close to outweighing negatives that come with generative AI. The fear of it replacing critical human jobs, the copyright concerns over AI training, the challenges its created for academic integrity and the decreased traffic to independent sites easily outweigh these benefits.

But it is important to be fair and to acknowledge that, in some ways, AI is actually improving things. Though many worry about the long term and short term impacts of AI, no one is mourning the struggles of essay mill sites or the decline in human-on-human plagiarism.

Simply put, authorship has become steadily more complicated over time. New ways to collaborate, new tools for editing and new processes for creation have consistently blurred the lines between author and contributor. Generative AI is just the final form of that challenge.

For better or worse, AI isn’t going anywhere. The question isn’t whether AI should be used, it’s how it should be used. To that end, there’s a broad spectrum when it comes to AI usage. There’s a major difference between using technology to correct errors and generative ideas versus using it to write the entire paper.

Thinking about and discussing these differences will be critical in all area. Whether you’re in academia, journalism, creative authorship or some other field, it’s important to not surrender human authorship while still finding ways to benefit from the technology.