T.He’s responsible for churning out some of the anger-inducing clickbait headlines crawling around your site Facebook feed They may not actually be people.In a report released Monday, researchers say They found 49 examples of news sites that generated articles. ChatGPT style AI chatbotThe articles identified share some common chatbot traits, but NewsGuard warned that “low-key readers” would never know they were written by software.
of The website spanned seven languages and covered subject From politics and technology to finance and celebrityAccording to a report from NewsGuard, a company that makes browser extensions that assess the credibility of news websites,None of the Sites recognized in the article they used artificial intelligence Generate stories. Regardless of subject matter, the website produced a ton of low-quality content and was littered with ads. Similar to human-generated digital media, this flood-the-zone approach aims to maximize potential advertising revenue. In some cases, one of his AI-powered websites pumped out hundreds of articles a day, some of which were clearly false.
“In short, a number of more powerful AI tools have been announced and made available to the public in recent months, so they could be used to rouse an entire press that has been the subject of speculation by media scholars. The concern that there is now a reality,” Newsgard said.
While most of the content NewsGuard reviewed appears to be relatively low-risk content farming aimed at generating easy clicks and ad revenue, some sites go a step further and potentially You are spreading dangerous misinformation. a site, celebritydeath.com, After posting an article claiming that President Joe Biden “died peacefully in his sleep,” Vice President Kamala Harris took over.
of The first line of the fake story about Joe Biden’s death is continued ChatGPT Error message: “Sorry, we are unable to complete this prompt as it violates OpenAI’s use case policy regarding the generation of misleading content. It is not ethical to fabricate news about the death of someone, especially a celebrity like the president. ”
IIt’s unclear if OpenAI’s ChatGPT played a role in all the site’s articles, but it’s certainly the most popular generative chatbot and the best known.OpenAI did not immediately respond to Gizmodo’s request for comment.
Chatbots have some dead giveaways
Many of the stories generated by AI are it is clear. Nearly all of the websites identified reportedly used robotic, soulless language. Time spent with AI chatbots has gotten used to it. In some cases, the bogus her website didn’t even bother to remove the wording where the AI explicitly reveals itself. A site called BestBudgetUSA.comfor example, According to the report, he published dozens of articles containing the phrase “I’m not capable of making 1500 words” before providing a link to the CNN article.. All 49 sites There was at least one article with an explicit AI error message like the one above, the report said.
Similar to human digital media, most of the articles identified by NewsGuard were abstracts of articles from other prominent news outlets like CNN. In other words, there are no detailed explainers or investigative reports here. He only admitted to using AI on two sites, according to NewsGuard.administrator of one site It said it might use AI to generate the content, but said its editors had properly fact-checked it before publishing.
Ready or Not, AI Writers Are Ready
NewsGuard reports are Concrete numbers showing growing interest among digital publishers in leveraging AI chatbots. Will readers actually embrace them? The Reality of AI Writers It remains far from certain,Earlier this year, technology news site CNET announced that indexOsed for Use ChatGPT style AI generating dozens of low-quality articlesmany of which are full of errors, without informing the reader.Not only boring, but AI-generated content written under the signature “CNET Money”” contained scattered factual inaccuracies.The publication eventually had to issue a major revision, and in the months that followed, how no Deploy AI-generated content.
Meanwhile, the CNET debacle didn’t stop other big publishers from tinkering with generative AI.Last month, Insider Global Editor-in-Chief Nicholas Carlson said, sent a note to the staff The company says it will create a working group to explore AI tools that can be incorporated into reporters’ workflows.Some journalists reportedly use generated AI for testing Use tools to create story text, use tools to create outlines, prepare interview questions, experiment with headlines, and more. Eventually, the company will reportedly roll out AI principles and best practices across the newsroom.
‘A tsunami is coming,’ Carlson said Axios“We are either going to ride it or be wiped out by it.”
