Pro-censorship group Newsguard is working to train AI models to fight misinformation, while rating Chinese state-owned media outlets as more reliable than American, center-right news outlets, according to a report.
The Washington Free Beacon reported that Newsguard, the pro-censorship group operating a blacklist against conservative media, announced a tool in 2023 for AI companies. Newsguard claims to rate the credibility of a news outlet based on the likeliness that the outlet would spread mis- or disinformation. The organization then sells its data to advertisers that would then discourage said advertisers from placing ads on disfavored media.
Breitbart News reported that GDI claimed that the top ten outlets at most risk of spreading disinformation included center-right outlets such as the Federalist and the New York Post, while those labeled as the least likely to spread disinformation included HuffPost and ProPublica.
Newsguard reportedly encouraged AI companies to use its data during the “supervised learning” process, a technique that influences how AI models analyze information and coordinates AI outputs with developer preferences. The Free Beacon reported that Newsguard would license its data in a “machine-readable” form that would be ideal to train AI chatbots to discourage the spread of disinformation.
“Without tools like NewsGuard for AI, which can be used during supervised learning to train transformer models, AI language models risk becoming … the ‘next great misinformation superspreader,’” Newsguard said in a press release.
Newsguard COO Matt Skibinski argues that Newsguard does not market its data “for AI model training.”
The Free Beacon explained that an AI model trained on Newsguard’s ratings may treat American news outlets as less reliable than Chinese state-owned media. The outlet cited Newsguard’s rating of the China Daily, a state-owned outlet, which received a 44.5 score out of 100. This is more than twice the reliability of Newsmax, which received a score of 20.
Newsguard said that China Daily “does not publish false or egregiously misleading content at a significant level. In contrast, Newsguard said that Newsmax “has advanced false or egregiously misleading claims on health, U.S. politics, and the Jan. 6, 2021, attack on the U.S. Capitol, such that a visitor to the site risks being seriously misled.”
Other center-right outlets such as the Federalist and One America News Network risk “being significantly misled on important topics.” Both outlets received a score of 17.5 and 22.2. China Global Television Network, which Newsguard said was a “mouthpiece of the Chinese Communist Party,” scored a 44.5 score, which is five points below the Daily Wire, which received a 49.5 score.
Mike Benz, a former State Department official and the executive director of the Foundation for Freedom Online (FFO), told the Free Beacon:
The obvious danger of NewsGuard’s entry into AI is that it induces the same censorious changes to AI that it did to media and social media. The fear every citizen should rightly have is that a highly biased private firm could influence or effectively control the sources from which AI derives its decisions, the same way NewsGuard’s blacklists were used to control the media sources to which audiences could have access.
Newsguard says it has a “suite of protections to safeguard AI productions from election misinformation. “In a year with more than 40 major elections worldwide,” a 2024 press release stated, adding that “the new tools will protect AI products from being exploited or causing inadvertent harm.”
The Free Beacon reported that it remains unclear which AI companies purchased data from Newsguard, which itself does not disclose its clients. However, Anthropic and Microsoft have said they do not use Newsguard data to train their respective models. OpenAI and Google Deepmind did not respond to the outlet’s requests for comment.
“It’s the same fundamental issue as before, but AI makes it even easier to silence dissenting views,” Gene Hamilton, the president of America First Legal, said. “Who decides what’s reliable and what’s not?”
Newsguard COO Skibinski said “AI companies do not license our data about domestic news sites.”
“The only AI-industry use of our data has been through our FailSafe for AI product,” he explained, referring to Newsguard data focused on Russian, Chinese, and Iranian sources.
“AI companies would be welcome to license our full dataset if they wished to do so,” he added.
The Newsguard COO disputed the notion that Newsguard gives lower ratings to center-right news outlets than Chinese state-run media.
He argued in a statement to the Free Beacon:
Your statement that NewsGuard gives ‘lower ratings to rightwing American outlets than to Chinese state-run media’ is not accurate, no, despite the cherry-picked set of sites you describe here,” Skibinski wrote in an email. “If what you’re instead asking is whether a particular website, Newsmax, had so much false or egregiously misleading content on its site that it somehow managed to score even lower than a Chinese state media outlet, that is true—but the fact that so many other conservative-leaning sites are rated credible is evidence enough that Newsmax’s score is not based on its political leaning.
“There’s no measuring stick for a NewsGuard to measure the veracity of a state-controlled news outlet in China,” Hamilton told the outlet, arguing Newsguard should not rate Chinese state-owned outlets at all. “You’re never going to get actual, unbiased information.”