Major global brands are sharing advertisements on artificial intelligence (AI) -generated websites through programmatic advertising, unknowingly funding these unreliable news and information sources.

Analysis by NewsGuard discovered a major rise in what they call unreliable artificial intelligence-generated news information and news sites (UAINs), which are being funded by automatically programmed ads by major brands.

UAINs are sites which operated mainly via AI technology, producing news stories and information with little human oversight, making them often unreliable sources.

According to NewsGaurd, the number of UAINs has risen from 49 to 217 in just the last month.

These sites are mainly funded through programmatic ads by major brands – these ads are automatically programmed to be places on websites fairly indescriminately.

As UAINs can produce hundreds of bot-generated content a day, these are prime targets for these programmatic ads as they would algorithmically appear to be attracting more traffic due to such a high content output.

Companies still have to pay every site to run their advertisements, so these programmatic advertisements are inadvertently funding these unreliable, AI-generated websites.

NewsGuard identified 393 programmatic ads from 141 major brands in May and Jun 2023 which appeared on 55 UAIN sites across the US, Germany, France, and Italy.

While NewsGuard did not disclose the specific brands that appeared on the sites, they did include a wide range of advertisers, including major banks and financial services firms, appliance manufacturers, consumer technology companies, e-commerce companies, broadband providers, and a supermarket chain in Europe.

Over 90% of the ads viewed by NewsGuard were served by Google Ads, the largest online ad platform. Programmatic advertising follows users with targeted ads, so rather than being website specific, they are user-oriented.

This means most companies were likely unaware their ads were funding these unreliable sites, as they often go through third parties to monitor their advertisements. Further, while many advertisers do have mitigates to not advertise on certain sites, these appear to have not kept up with the rising number of UAIN websites.

The majority of UAIN sites were low-quality, according to NewsGuard, but did not contain misinformation or plagiarism. Still, some appeared to copy and rewrite articles from reputable news sites, and others appears to make unproven health claims.

Google’s ad policies actually state that their placement service will not allocate ads to “automatically-generated content.” Despite this, 90% of the ads NewsGuard found on UAIN sites were served by GoogleAds.

Further, Google places no limitation on the number of ad widgets – the pieces of code where developers can place their programmatic ads – on websites, meaning these UAIN sites can plaster ads across their pages, generating more money.

Recommended

• Is “Freedom of Choice” Impacting Tech Subject Uptake at Schools?
• Will AI Change the Workforce Structure? It Already Has
• Report: CX and EX are Key Drivers in Cloud and AI Adoption

On one site, an AI-generated article has 15 accompanying Google Ad widgets, some of which were from reputable brands.

Google exults the ease of their programmatic ad service – all programmers have to do is to use their Google Adsense code in their website, and it will begin to work, filling with programmatic ads which they will then earn money from.

UAIN sites are likely more prolific than NewsGuard is even able to identify – analysis could only discern if a site was AI-generated if they had an AI-language model error message on their site.

There’s likely more out there that have remained unidentified. NewsGuard did reach out for comment from both the UAIN sites and the brands which had ads appearing on the sites – they received either no response or a decline for comments.