Discover more from The Shortcut
CNET's AI writer appears to be stealing content
The AI engine is very good as plagiarizing content, it seems
The situation around CNET’s controversial use of artificial intelligence to generate articles has gone from bad to worse, as the suggestion has been made that the AI engine plagiarized content from human journalists.
CNET owned up to using AI for content generation a couple of weeks ago, after it was found that several financial explainer articles on the site were surreptitiously authored by “automation technology”. The Internet did not respond well, and despite assurances that every AI-generated post was “reviewed, fact-checked and edited by an editor with topical expertise”, the appearance of several basic math errors in the AI’s published work only sparked more backlash.
➡️ The Shortcut Skinny: CNET’s AI woes
😶 CNET’s AI appears to be plagiarizing content from other sites
👎 A new report details several instances of lazy copying
🤦♀️ The AI takes an article, jumbles a few words about, and passes it off as its own
😡 It’s not a good look for the already embattled CNET leadership team
Afterward, CNET’s leadership issued a hefty correction in a footnote and announced in an internal meeting that it would be pausing the publication of stories created using AI “for now”.
That’s just as well because it looks like the AI was becoming a little too enthusiastic with its digital pen. After combing through CNET’s AI articles, Futurism has found several instances of plagiarism across its body copy and headlines.
One example it highlights concerns a CNET article about overdraft protection. A sentence reads: “Overdraft fees and NSF fees don't have to be a common consequence. There are a few steps you can take to avoid them”. Which looks lovely and fine. Except it bears a striking resemblance to this other sentence published in a Forbes article about overdraft fees: “Overdraft and NSF fees need not be the norm. There are several tools at your disposal to avoid them”.
There are several other, similar examples in which CNET’s AI appears to have copied another site’s existing article, run every other word through a thesaurus, jiggled around the sentence structure a bit, and published the result. That’s not really writing, and it’s not really serving the reader.
"If a student presented the equivalent of what CNET has produced for an assignment in my class, and if they did not cite their sources, then I would definitely count it as plagiarism," Antony Aumann, a philosophy professor at Northern Michigan University, told Futurism.
"Now, there is some dispute among academics about exactly what plagiarism is. Some scholars consider it a form of stealing; other scholars regard it as a kind of lying. I think of it in the latter way. Plagiarism involves representing something as your own that is in fact not your own. And that appears to be what CNET is doing."
For all the bad press that’s surrounded the practical implementation of artificial intelligence in recent weeks, the tech isn’t going anywhere. After laying off 10,000 employees, Microsoft announced a multi-billion dollar investment in OpenAI, the company behind popular tools like ChatGPT and DALL-E, and said it plans to integrate similar AI models into its consumer products down the line. We’ll be encountering more AI in the future, whether we know it or not.
The Shortcut is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.