AI Fake News Sites Manipulating Google Ads – Investigation Reveals
AI-Generated fake News Farms Exploiting Google’s ‘Discover’ for Profit
A disturbing trend is rapidly polluting the digital landscape: the proliferation of thousands of fake news websites created using artificial intelligence, specifically designed to game Google’s search algorithms and generate significant advertising revenue. French investigative journalist Jean-Marc Manach has uncovered a network exceeding 4,000 such sites, primarily targeting Google’s “Discover” service.
Manach’s investigation, which began in early 2024 and expanded in collaboration with the newspaper Libération, reveals a sophisticated operation driven by SEO experts seeking to exploit loopholes in Google’s content suggestion system. These sites, many based in France with a growing number in English, rely on entirely copied or fabricated articles, prioritizing backlinks and visibility in the Discover feed – a personalized news section for Android users.
The scale of the operation is staggering. By October, Manach and his journalism students had identified 250 sites, a number that ballooned to over 4,000 through continued investigation. The financial incentives are important, with at least two site owners reportedly becoming millionaires through advertising revenue despite publishing demonstrably false content.
Google Discover: An “ATM” for Fake News
The core of the problem lies with Google Discover, which recommends content based on user interests. manach argues the system struggles to differentiate between legitimate news sources and AI-generated fabrications, effectively presenting these deceptive sites as trustworthy. Sources describe discover as an “ATM,” capable of generating thousands of dollars in advertising revenue daily, and a major driver of traffic, even surpassing traditional search.
the content itself is often sensational and designed to maximize clicks. Fabricated stories circulating on these sites include false claims about the cancellation of banknotes in France, restrictions on inheritance transfers, and the misappropriation of french savings to fund the war in Ukraine. More outlandish tales, such as the discovery of a 25,000-year-old pyramid or a giant predator beneath the Antarctic ice, also proliferate.
Crucially, these articles frequently feature AI-generated images and addresses, meticulously crafted to provoke curiosity, fear, and ultimately, clicks.
The rise of ‘Obstetric AI’ and SEO Tactics
Manach traces the surge in these sites to instructional videos on YouTube from 2023 and 2024 detailing how to “penetrate” Discover using “obstetric AI” – a term referring to techniques for rapidly generating content optimized for the algorithm. The French market has become saturated, prompting some operators to target foreign language markets.
The investigation has identified over 120 companies and individuals behind these sites, including journalists and media trainers. Alarmingly, some sites have replaced human journalists with AI without disclosing this to the public.
Google’s Response and Calls for Accountability
Google acknowledges the issue and states it enforces strict policies against unwanted content, regardless of its origin. The company claims its systems actively demote low-quality content in Discover and search results.
Though, Manach dismisses these measures as insufficient. He advocates for a clear distinction between human-authored and AI-generated content, urging journalists to authenticate their work and demanding clarity for the public. He calls for accountability for those who conceal the use of AI in content creation.
“This phenomenon doesn’t just threaten the press, it threatens democracy and the public’s right to reliable facts,” Manach emphasizes. He hopes his investigation will inspire further journalistic efforts to document these sites and combat the growing ”information pollution” in the digital space. The need for vigilance and proactive measures to safeguard the integrity of online information has never been more critical.