Fake media publishes 1,000 articles a day: AI reaches content farms | technology

When a web page publishes articles in a massive way, seeking income by weight from ads, it is called a content farm. They’ve been around for years, perhaps ever since someone saw a business in creating content at a low cost, the better, to make it profitable through ads served through automated platforms, such as Google AdSense. With artificial intelligence for text generation, such as ChatGPT or Google Bard, this dynamic takes on an industrial scale.

A study by disinformation monitoring platform NewsGuard confirms that there are more and more content farms that use generative AI. These sites feed on articles generated by chatbots and apparently lack editorial support. His numbers are astronomical. In the week of June 9, one of the sites examined, world-today-news.com, posted about 8,600 articles, an average of 1,200 per day. Two more pages on NewsGuard’s Radar had 6,108 posts and 5,867 posts during that week.

What is clear is that they use artificial intelligence to create content clickbait Low quality,” says Mackenzie Sadeghi, senior analyst at NewsGuard. “Websites use this technology to produce articles faster but also cheaper.” Sadeghi points to the fact that people practically disappear from the equation: “Before, these web pages had a team of human collaborators, Freelance work For those who are paid to write content. Now there doesn’t seem to be much human oversight,” adds the expert.

This is why these content farms post chatbot error messages as if they were news headlines.And These clues guided researchers to discover written content using artificial intelligence. In their research, NewsGuard analysts came across statements like “Sorry, I’m an AI language model, I can’t access external links or web pages on my own”; Or other annoying messages: “I’m sorry, I can’t abide by these instructions because it goes against my moral and ethical principles.” The fact that no one is there to delete these messages indicates the degree to which these sites are automated.

See also  Metaverso, Year One: A Dangerous and Millionaire Bet to Retain Young People Denying Social Networks | technology

Content farms are incentivized to create articles at their discretion. The more posts they make, the more visitors they will attract to their websites and the more users will click on the ads they have. NewsGuard has verified that more than 90% of this advertising is served with Google ads, whose algorithm automatically places ads on the pages attached to the platform. Advertisers also voluntarily sign up to this advertising platform and their messages may end up on these content farms. Between May and June, the analysts identified 393 ads from 141 highly relevant brands on 55 highly relevant web pages.

Google does not have a policy banning AI-generated content, but it does have a policy banning content that tends to Spam emails And of low quality, which is mainly what these pages offer,” Sadeghi comments. The American giant earned $224.470 million from ads in 2022, according to Statista data. Although only a small part of that number is attributed to automatic ads, since most of Billing comes from search ads.

Error message from a generative AI deployed to a content farm without human supervision.
Error message from a generative AI deployed to a content farm without human supervision.

The use of generative AI in content farms is rapidly increasing. “We are discovering between 25 and 50 such sites per week. At the beginning of May we identified 49 and we now have 277 on the list. Some are new, some have been around for years and they are now starting to use AI,” says the senior analyst.

Most of the sites NewsGuard analyzes, with advertisements from relevant brands, do not spread false news. Sometimes they get into the realm of misinformation, with titles like Can lemon treat skin allergy? also Five natural remedies for attention deficit disorder. But, in general, they can only be blamed for their low quality, often with stolen content.

The real problem comes from the combination of generative AI and misinformation. In Spain, CSIC researcher David Arroyo, who is working on detection False newsassociates artificial intelligence with a greater ability to create fake news: “The phenomenon of disinformation will undoubtedly increase due to the fact that these tools exist,” he emphatically asserts.

ammunition for disinformation

Journal article nature I already warned, back in 2017, about the association between fake news and auto ads. He argued that most of the fake news created during the 2016 US election campaign was not politically motivated, but economically motivated. “There has already been talk of a whole ecosystem of advertisers involved in creating domains for invented content and distributing it. With these AI tools, everything is amplified, because the ability to artificially create authoritative content has increased tremendously,” Arroyo explains.

They find out from CSIC that disinformation has increased in recent months, although Arroyo does not attribute everything to artificial intelligence: “It would be difficult to isolate a single cause. It must be taken into account that in Spain we have been in electoral processes since May, and added to It has all the elements of distortion of Russian movements related to the war in Ukraine,” he adds.

A few months ago, NewsGuard ran a study on ChatGPT, in versions 3.5 and 4, to assess its potential as a fake news inventor. Chat bots Sadeghi asserts that they were able to create disinformation about topics such as politics, health, climate or international issues. “The fact that they are able to produce such disinformation, when someone directs them, shows how easy it is to manipulate these models’ defenses,” says the NewsGuard analyst. Added to this is its amazing ability to produce content in an industrial way.

You can follow The Country Technology in Facebook And Twitter Or sign up here to receive The weekly newsletter.

Subscribe to continue reading

Read without limits

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button