Days after a report surfaced pointing out that a trove of articles from CNET labeled as written by “staff” were actually scribed by an artificial intelligence (AI) tool, the fabled site is in the crosshairs of yet another controversy.
A new investigation conducted from Futurism uncovered “extensive evidence” that suggests the AI-generated articles are so close to work of competitors like Forbes that they constitute plagiarism. Finding “deep structural and phrasing similarities” and providing side-by-side examples, Futurism asserted that “the bot directly plagiarized the work of Red Ventures competitors, as well as human writers at Bankrate and even CNET itself.”
TheWrap previously reported on CNET attempt’s to right course by removing “staff” from the bylines of AI-written articles, after they were brought into question in an earlier Futurism report. Then, the publication found the content was “riddled with factual errors.” Now the site’s report likens CNET’s AI work to that of a “a serial plagiarist.”
TheWrap reached out to CNET for comment but hasn’t yet heard back. CNET included disclosures on pages featuring AI-generated articles and noted they were scrutinized by a human editor before being published.
In its report, Futurism provided side-by-side examples of the cited plagiarism. The site reported that CNET’s AI parses already-published articles and “makes adjustments — sometimes minor, sometimes major — to the original sentence’s syntax, word choice, and structure. Sometimes it mashes two sentences together, or breaks one apart, or assembles chunks into new Frankensentences. Then it seems to repeat the process until it’s cooked up an entire article.”
The process reeks of classic plagiarism found in K-12 and higher education. To preempt the problem, a number of school systems, from New York City’s public schools to Seattle schools, have barred students from using the AI chat bot ChatGPT from OpenAI, a Microsoft-backed firm that exploded in popularity over the last three months with first-to-market mover advantage.
How CNET continues to engage with AI-generated content moving forward remains to be seen, but so far the publication’s top brass aren’t exactly forthcoming with details such as “what data had been used to train the AI.” Futurism cited a report in the The Verge about a CNET meeting in which the executive vice president of content and audience “refused” to elaborate or provide details.
