DigitalMANIPULATIVE

AI-Generated News Sites

What it is

Websites dressed as news outlets whose articles are produced mostly or entirely by generative models with little or no human oversight, built to harvest advertising revenue or to launder narratives under a newsroom-like name.

How it works

Readers judge a source in seconds from surface cues: a masthead, a plausible name such as a city plus Times or Gazette, a byline, a grid of headlines. Generative tools make those cues nearly free. NewsGuard began counting such sites in May 2023, when it found 49; its tracking center listed more than a thousand within a year and, by its June 2026 update, several thousand AI content farm sites in 16 languages. Most exist for programmatic advertising, which places brand ads automatically and so pays for pages no human edited. A smaller set is political: networks of fake local outlets that mix rewritten wire copy with planted stories. The harm is partly direct (fabricated stories, including false death reports) and partly ambient: search results and chatbot answers fill with text that cites nothing and corrects nothing. The count comes from one commercial rater, so treat it as an indicator, not a census.

Real-world examples

  • •NewsGuard's AI Tracking Center, launched in May 2023 with 49 sites, defines an unreliable AI-generated news site as one produced mostly or entirely by AI with little to no human oversight. Its running total had passed 3,700 sites by mid-2026.
  • •NewsGuard analysts have repeatedly found articles on such sites that still contain chatbot error messages, such as text stating that the request cannot be completed as an AI language model, published as if they were headlines or body copy.
  • •OpenAI reported in May 2024 that people acting for the Iranian operation International Union of Virtual Media used its models to generate and proofread long articles and website tags, one of which was published with a model message still embedded.
  • •In April 2024 Microsoft reported that the China-linked actor it tracks as Storm-1376 used AI-generated news anchors and audio in content aimed at Taiwan's January 2024 election.

Ethical guidelines

  • ●Using AI tools inside a newsroom with editors, corrections, and disclosure is a production choice. Presenting unedited machine output as reporting by named journalists who do not exist is deception.
  • ●Advertisers and ad-tech intermediaries fund most of these sites without choosing to; brands have a responsibility to know where their programmatic spending lands.
  • ●Naming a site as AI-generated is a factual accusation and should rest on evidence such as leftover model messages, not on a dislike of its politics.

How to defend against it

  • ►Leave the site to evaluate it (lateral reading, Wineburg & McGrew; the SIFT method, Caulfield). Search the outlet's name plus the word about or owner. A real newsroom has a traceable address, named staff with histories elsewhere, and a corrections policy.
  • ►Check the byline. Search the author's name; invented reporters have no other footprint, and their headshots may be synthetic.
  • ►For a striking story, look for the original. Content farms rewrite other outlets; if no primary source, document, or named witness is linked, wait for a second independent report.
  • ►Be especially careful with unfamiliar local-sounding names during elections; fake local outlets trade on the high trust people place in local news.
  • ►When a chatbot or search summary cites a source you do not recognize, open it and apply the same checks. The summary inherits the quality of what it read.

References

  1. NewsGuard (2026). Tracking AI-enabled Misinformation: AI Content Farm Sites (and Counting). NewsGuard AI Tracking Center, continuously updated; accessed September 2026 · link
    The definition of unreliable AI-generated news sites, the May 2023 starting count of 49, and the running total across 16 languages.
  2. OpenAI (2024). AI and Covert Influence Operations: Latest Trends. OpenAI threat intelligence report, May 30, 2024 · link
    IUVM generating and proofreading articles and website tags, including a published tag exposing AI generation.
  3. Microsoft Threat Analysis Center (2024). Same targets, new playbooks: East Asia threat actors employ unique methods. Microsoft Security Insider, April 2024 · link
    Storm-1376 use of AI-generated anchors, audio, and memes around Taiwan's January 2024 election.
Last reviewed
Suggest a correction

Detect AI-Generated News Sites in any text

Paste any message, email, or article into our free Manipulation Detector to see if AI-Generated News Sites or other techniques are being used on you.

Try the Detector FreeCreate Free Account

Related Articles

Dark Patterns in UX: How Apps Manipulate Your Behavior

Subscription traps, misleading interfaces, and engineered addiction. Understanding the persuasion techniques built into the apps you use every day.

7 min read

OSINT for Beginners: Open Source Intelligence Explained

Open Source Intelligence (OSINT) uses publicly available data to gather actionable insights. Here is a beginner-friendly guide to what OSINT is and how it is used.

10 min read