The sports news story you clicked on could be an AI flop
NBC Sportz did not respond to requests for comment. Neither NBCSport.co.uk nor BBCSportss.co.uk have an email address or other publicly associated contact information, so WIRED had no way of getting in touch. (All three sites were registered by the domain management company Namecheap, as was a site impersonating CBS News that DoubleVerify suspects is part of the Synthetic Echo network.)
Bad actors have attempted to poach successful media companies by republishing their work without permission many years. Now, however, AI tools are allowing variations of this scheme to spread at an even faster pace. “This type of low-quality content isn’t really new,” says Saporta. “But with these current tools, it’s much easier to replicate and scale.”
The number of AI slop websites has increased significantly year-over-year since generative AI tools became increasingly popular in 2023. Last February, shortly after WIRED first began reporting on the rise of AI content mills, the media monitoring company released NewsGuard had identified 725 “news and information sites” filled with AI content. By January 2025 the time had come identified at least 1,150 of these sites.
“The volume has increased,” says Shouvik Paul, chief operations officer at AI detection company Copyleaks. “Many of them are run by foreigners and are very shady operations. So how can you even keep up?”
To make things even more confusing for readers, some mainstream media sites have done this experimented with the publication of AI-generated news articles. (Sports Illustrated itself published supposedly AI-generated content that its parent company said was provided by a third party.) In other cases Domain Hitler have purchased the URLs of media properties that have fallen on hard times and brought her back to life as AI content mills that sometimes replace their previously well-founded journalism with robotic pablum.
Some of these websites are already causing confusion in the real world. an SEO content mill in October posted an AI-generated announcement for a Halloween parade in Dublin, Ireland. Even though no such event was planned, throngs of revelers arrived in anticipation of the festivities.
Paul from Copyleaks described the way some of these sites used the brand identity of real vendors to sell junk as “something like phishing.” In some cases, these websites actually appear to be conducting phishing attacks. One of the websites identified by DoubleVerify within the ring was designed to impersonate a Fox news outlet based in Nigeria. It greets potential readers with a series of suspicious software pop-up ads.
While the pop-ups look fake, the sites in this group appear to be doing a brisk business in programmatic ads, which are ads served via large-scale automated ad buys, rather than a direct relationship between specific sites and advertisers. Many feature a plethora of banners managed by popular programmatic ad servers like Criteo and Sharethrough. (Neither Criteo nor Sharethrough responded to requests for comment.) DoubleVerify’s report suggests that Synthetic Echo’s operators chose sport as one of the main content categories, particularly because it is considered more brand-safe than hard news.
As WIRED monitored these sites, programmatic ads appeared from a number of well-known companies, including tech giants like Asana and Oracle, e-commerce giant Net-A-Porter, makeup giant Sephora, and resort chain Kalahari Resorts. None of these companies responded to requests for comment.
At a time when trust in the media has plummeted and many news outlets are seeing revenue decline, this kind of negligent content mill ring is a double whammy. It pollutes the information ecosystem with garbage and stolen texts and siphons programmatic advertising revenue from legitimate content producers.
