Researchers: as social networks clamp down on disinformation, Russia has begun spreading propaganda in comments sections on news outlets, such as Daily Mail
Online sites test new methods as social media platforms try to scrub their platforms of propaganda Tweets: @rasmus_kleis , @campuscodi , and @1br0wn See also Mediagazer Tweets: Rasmus Kleis Nielsen / @rasmus_kleis : “You can't predict when Facebook will decide to take something down,” @vassgatov says, and argues Russian disinformation and propaganda campaigns have “switched from social networks to other channels of communication.” https://giftarticle.ft.com/... Catalin Cimpanu / @campuscodi : If I were a respectable news outlet, I would have disabled my comments section half a decade ago. https://twitter.com/... @1br0wn : “comment sections of western media outlets such as the Daily Mail, the Daily Express, Fox News and Der Spiegel are increasingly being manipulated by propagandists, who then point to pro-Kremlin comments as evidence of sympathy for the Russian government” https://www.ft.com/... See also Mediagazer
Context & Ripple Effects
This report is the earliest data point in a channel-shifting arc the corpus keeps documenting. After Meta began fact-labeling four Russian state media outlets and refused Moscow's order to stop (Nick Clegg said Meta refused a Russian order to halt fact-checking), then globally demoted links to state media on Facebook and Instagram (Meta's global demotion of Russian state media links) while Russia itself restricted access to Twitter and Facebook at home, researchers observed propaganda migrating off social networks entirely — into the comments sections of Western outlets like the Daily Mail, Daily Express, Fox News and Der Spiegel.
What makes it matter is what came after: the migration didn't end with comments sections. An [[a:843809|EU study found Kremlin-backed accounts grew in reach through 2023 despite platform measures]], and by late 2025 the same network was flooding the web with 'LLM grooming' content aimed at chatbots. Each enforcement point on the big platforms has so far displaced the campaign rather than contained it.
First-order effects
- Publishers like the Daily Mail and Fox News inherit a moderation burden they never signed up for: their comment sections are now an active distribution channel for pro-Kremlin sentiment, forcing them to choose between staffing up moderation or shutting comments — Catalin Cimpanu's own prescription was that respectable outlets should have disabled comments years ago.
- The named outlets' brand-safety calculus changes immediately, since advertisers and readers now associate their properties with manipulated discussion threads rather than neutral audience chatter.
Second-order effects
- Platform crackdowns function as displacement, not deterrence: Meta's labeling and demotion of state media pushed the operation onto unmoderated surfaces, meaning every new enforcement rule on one channel raises the value of the next undefended one.
- Lawmakers who pressured Meta and Twitter over Russian embassy accounts get a new target class — news publishers — extending the moderation debate from platforms with clear policies to outlets where comment moderation has no regulatory framework at all.
Third-order effects
- If the pattern holds, state information operations become structure-seeking: they route around whatever surface acquires a moderation apparatus, which means containment requires monitoring the whole open web rather than negotiating with a handful of platforms — the trajectory the corpus traces from comment sections to the Pravda network's later grooming of AI systems.
- The long-run effect is asymmetric defense economics: platforms can be legislated and publishers can close comments, but each closure cedes ground to whoever is willing to operate on ungoverned terrain, steadily shifting influence operations toward channels no single regulator controls.
The trend: State-aligned propaganda is systematically migrating away from moderated social platforms toward whatever surface lacks a moderation regime — first publisher comment sections, eventually the training corpora of AI systems.