A new report from Model Republic publication The Midas Project found that the news site, The Wire by Acutus, relies almost entirely on AI-generated content. This publication has been active since the end of 2025 with almost 100 articles published in all areas of technology, energy, media, science, business, and healthcare. Unknown yet, theirs About the page describes their work as “participatory journalism” led by an “editorial team,” but the site has no masthead and states that it has no editors or journalists in its publications.
The official explanation for this anonymity is buried in their How It Works article:
Our editorial team identifies timely topics and invites well-informed, first-hand contributors to share their perspective through structured discussions. Those ideas are combined and organized into stories that show where the participants agree, where they diverge, and what it all means – providing depth, balance, and clarity beyond the subject.
But while reporter Tyler Johnston runs the site’s content Pangraman AI detection tool that boasts an accuracy rate of 99.98%, found out how widely AI was relied on: “Of these 94 articles, 69% were also marked as fully AI-generated, and another 28% were marked as partially AI-generated. Only three articles were defined as human-authored.”
Johnston’s suspicions grew when he looked at the content itself, which was overwhelmingly in favor of the development of artificial intelligence and dismissive of AI critics. One passage, for example, warns “The Rise of Anti-AI Radicalism,” while another criticizes the reader: “Will Republicans Let Green States Set America’s AI Rules?”
Mashable Light Speed
As Johnston dug deeper, the picture became clearer. As a new site with very little social media, The Wire’s articles rarely get retweeted, but Johnston found that part of his connection to X came from Patrick Hynes, president of PR firm Novus Public Affairs. A quick look their customer list reveals that they work for Targeted Victory, a consulting firm involved in OpenAI’s lobbying efforts in Washington on behalf of its regulatory interests.
Productive artificial intelligence is already creating tension in our collective perception of reality. With enough computing power, you can create fake trailers to films that have never been made and never will be, either steal the political voice for deeper deceptionor create an absurd, unbelievable situation, like this a shark attacking a planeand fool at least a few gullible internet rookies.
If Johnston’s reporting is true and his views are accurate, we may have a case of an AI company deliberately mislabeling its work as “independent journalism” to justify itself (something Johnston is suggesting contradicts his own). usage policies).
Disclosure: Ziff Davis, Mashable’s parent company, in April filed a lawsuit against OpenAI, alleging that it infringes Ziff Davis’ copyrights in training and using its AI programs.
Articles
Artificial intelligence