Last November, sensational news about the Israeli prime minister’s psychiatrist committing suicide went viral on the internet. However, the news was generated by artificial intelligence (AI) originating from one of hundreds of websites, which researchers say produces technology-based ‘fiction’ masquerading as news.
Websites spreading propaganda used to rely on armies of buzzers, but generative artificial intelligence tools now offer a much cheaper and faster way to create content that is often difficult to distinguish from authentic information.
Hundreds of AI-powered sites mimicking news sites have emerged in recent months and fueled an explosion of false narratives, from wars to politicians, that researchers say is fueling concern in a high-stakes election year across the world.
“The news ‘The Israeli Prime Minister’s psychiatrist committed suicide’, still tops the list of popular articles highlighted on Global Village Spacea Pakistani digital site, after they caused a stir last November with baseless claims blaming Netanyahu for the incident.
Much of the site’s content, including the article, appears to be pulled from mainstream sources using AI tools, according to the analysis NewsGuard, a US-based research organization that tracks misinformation.
After scanning the site for error messages specific to the content generated by the AI chatbot, NewsGuard said it found significant similarities between the story about Netanyahu’s “psychiatrist” and a fictional 2010 article on a website that ran satirical news stories.
Also read: Elon Musk sues OpenAI for allegedly betraying its founding mission
Analyst NewsGuardMcKenzie Sadeghi said when he asked ChatGPT, from Microsoft-backed OpenAI, to rewrite the original article for a general news audience, the results were very similar to the article in Global Village Space.
“The exponential growth in AI-generated news and information sources is particularly worrying because these sites may be perceived by the average user as legitimate and trustworthy sources of information,” Sadeghi told AFP.
Pushing propaganda
Also read: Presidential Decree on Publisher Rights Ratified, There is a 6 Month Transition Period
The fake article, which appeared as Netanyahu called for war against Hamas militants in the Gaza Strip, also spread across social media platforms in multiple languages, including Arabic, Farsi and French.
A number of sites published obituaries of these fictional “psychiatrists.” This fake news was also featured on a television show in Iran, Israel’s archenemy, when the host directed viewers to read the full article at Global Village Space.
Also read: OpenAI Reportedly received IDR 1250 Trillion Fund Injection
The website, which rebranded Netanyahu’s article as “satire” after criticism, did not respond to AFP’s request for comment.
NewsGuard has identified at least 739 AI-generated “news” sites in multiple languages that operate with little or no human oversight and come under generic names like “Ireland Top News,” and so on.
“But the list may just be something that’s easy to find,” said Darren Linvill, of Clemson University.
Also read: OpenAI in Agreement with Investors Worth US$80 Billion
He is a disinformation expert at the university who discovered several Russia-linked websites were spoofing news and pushing Moscow’s propaganda about the war in Ukraine ahead of the US presidential election in November.
They are included DC Weeklywhich follows NewsGuard using AI to rewrite articles from other sources without credit.
This site, which appears to be owned by John Mark Dougan, a former US marine who fled to Russia, has published a lot of false info including that Ukrainian President Volodymyr Zelensky bought two luxury yachts worth millions of dollars with American aid money.
Also read: Press Council Encourages Professionalism of Journalists to Guard the 2024 Election
The impact of this false information has influenced policy decisions. Some US lawmakers, for example, put forward a false narrative in the midst of an important debate about aid to Ukraine, because of this information.
“Automatically generated misinformation will likely be a major part of the 2024 election. Fraudsters are using (Generative) AI to influence everything,” New York University professor Gary Marcus told AFP.
“AI-generated content that populates websites like DC Weekly helps create a kind of camouflage that gives greater credibility to their false stories written by humans,” Linvill told AFP.
Also read: OpenAI Releases Guidelines for Measuring Artificial Intelligence Disaster Risk
The emergence of these websites underscores the potential of artificial intelligence tools, chatbots even more so than photo generators and voice cloners, to increase misinformation while further eroding trust in conventional media, researchers say.
Their polarizing content, which can incite chaos and influence political beliefs, is intended to attract attention and generate advertising revenue.
The revenue model for most of these websites is programmatic advertising, meaning that big brands may inadvertently endorse the site. “Meanwhile governments may find it difficult to take tough action for fear of violating free speech protections,” the researchers said.
“If we don’t stop it, it will only further erode the boundaries between reality and fiction which are already so blurry,” stressed Linvill. (AFP/M-3)
#Beware #Fake #News #Sites #Supported #Artificial #Intelligence #Increasingly #Rampant