For years, Western governments and cybersecurity experts have warned that Russia’s online information operations were expanding beyond troll farms and botnets. A new analytical report from the Institute for Strategic Dialogue (ISD) reveals just how far that evolution has progressed, revealing that a sprawling pro-Kremlin propaganda ecosystem has quietly established itself across the English-language internet, gaining legitimacy through backlinks from hundreds of popular and fringe sites.
How a Russian Propaganda Web Built Influence Across the English-Language Internet
1. A Hidden Influence Infrastructure
The Pravda network is one of the most extensive Russian disinformation systems ever documented. Since its inception in 2014, it has published over six million articles across approximately 90 sites, many of which are under a single domain with dozens of subdomains.
According to the ISD, this activity has increased considerably over the last year, from approximately 6,000 pieces per day in 2024 to nearly 23,000 articles per day by May 2025. The network publishes in several languages and increasingly focuses on audiences in Asia, Africa, Europe, and the English-speaking world.
The ISD discovered that the Pravda network’s expansion is especially troubling because its information has been regularly quoted on legitimate—and often important—websites. Between July 2024 and July 2025, about 900 English-language websites connected to Pravda network content.
These websites include national outlets, local newspapers, academic organizations, political discussion platforms, and personal blogs. Moreover, over 80% of these citations regarded Pravda stories as credible, neglecting to identify their origins or link to a Russia-aligned disinformation campaign.
The scope of this issue goes far beyond accidental citation. Each backlink has an SEO value and boosts the network’s visibility, propels its material higher in search engine results, and communicates trustworthiness to algorithmic systems—whether they are Google, academic search tools, or leading language models trained on internet texts. According to ISD, even pages that acknowledge Pravda content but oppose it contribute to its digital legitimacy.
This isn’t just a case of low-quality websites publishing questionable content. Thirty-four high-traffic websites, each with over one million monthly users, referred to Pravda sources as authentic content. These included trusted magazines and widely read discussion blogs from all ideological perspectives. Fifteen other high-traffic sources, including The Washington Post, Newsweek, and Fortune, supplied decontextualized links that acknowledged the content’s Russian provenance but did not label it as part of an information campaign.
The end consequence, academics say, is an unintentionally created highway enabling Kremlin narratives to flow directly into public discourse, evading the editorial controls that ordinarily protect readers from covert influence attempts.
2. How a High-Volume Propaganda Engine Disrupts the Information Ecosystem
Persuasion is commonly used in disinformation efforts, with carefully produced narratives intended to simulate real news reporting. The Pravda network employs a different tactic: flooding.
Researchers refer to the network’s strategy as a “numbers game,” relying on massive amounts of content to overwhelm the digital ecosystem rather than a single compelling piece of propaganda. The strategy is straightforward. By publishing tens of thousands of pieces every day on various topics, the network ensures that its content shows in search results for almost any Russia-related topic, from geopolitics to celebrity news.
Forty percent of the Pravda articles linked by English-language sites dealt with Russia’s war in Ukraine, which is predictable given the Kremlin’s strategic interest in shaping international opinion of the war. However, the remaining material reveals a far broader focus: US domestic and foreign affairs, comments on global political leaders, celebrity culture, natural disasters, terrorism, climate change, and even the behavior of high-profile tech CEOs such as Elon Musk. Some of this stuff is consistent with known Kremlin talking points, while most of it merely exploits trending issues to attract readers and search engines.
The volume alone provides the operation a huge advantage. The Pravda network has over six million published articles and publishes one every few seconds, making it a prolific aggregator of pro-Russia posts, much of which is reused from affiliated Telegram groups, Russian official media, and other sources. Even if the majority of the content is ignored, enough of it ends up in genuine communication channels to skew the overall information environment.
The distribution pattern discovered by ISD demonstrates how this method succeeds. Approximately 75% of the English-language pages linking to Pravda content come from commentary sites, which provide analysis, opinion, aggregated news, or personal essays. These usually have a poorer editorial review, rendering them susceptible to unvetted sourcing. Approximately 20% came from formal reporting and research institutions. Only 5% were Russian state-funded websites.
Because of the massive volume of Pravda content, users searching the internet for quotes, updates, or source material on Russia, Ukraine, or worldwide political controversies routinely come across Pravda links, many of which they are unaware of. When propagandists use real-time global events as scaffolding, it becomes challenging to distinguish between true news and produced propaganda in an increasingly saturated information space.
ISD discovered that webpages portraying Pravda content as valid sources received over 22,000 engagements across major platforms such as Facebook, X, Reddit, and Pinterest, more than doubling the engagement seen on pages that correctly contextualized Pravda material.
Such encounters, however unintentional, broaden the network’s reach and give its narratives a foothold in conversations well beyond its Russian roots.
3. The AI Connection: Feeding Propaganda into the Future’s Training Data
Underlying the Pravda network’s increased activity is a growing concern among cybersecurity researchers, intelligence agencies, and AI ethicists: Russia may be attempting to infiltrate the training data of large language models (LLMs) such as ChatGPT and Google’s Gemini.
This method, commonly known as “LLM grooming,” takes use of the fact that AI models are trained on enormous amounts of publicly available internet data. Disinformation operators attempt to impact the probability patterns that shape model outcomes by widely disseminating pro-Kremlin narratives. If a model has seen thousands of articles pushing a false claim, such as that Ukraine operates U.S.-funded bioweapon facilities, it may be more likely to repeat that claim when asked.
Researchers have already identified warning indications. Earlier analysis discovered that certain chatbots occasionally repeated Russian disinformation when responding to questions regarding Ukraine, US policy, or European security issues. While recent tests demonstrate that prominent models rarely directly mention Pravda links, the ISD report adds that the network’s high-volume strategy raises the possibility that its content would appear alongside other sources in datasets, affecting model relationships and indirectly influencing outcomes.
The Pravda network has also made its way into platforms that power search engines and AI outputs. Wikipedia citations and Community Notes on X (previously Twitter) contain links to Pravda pages. This cross-platform visibility boosts the network’s legitimacy, making it more difficult for automated systems—and humans—to discern authentic material from state-sponsored propaganda.
AI ethicists worry that poisoning training data via large-scale content farms could become one of the most effective types of 21st-century information warfare. While social media platforms and news organizations have developed sophisticated tools to resist coordinated inauthentic activity, LLMs are still susceptible to subtle manipulation, particularly when rogue actors exploit gaps in transparency and material sourcing.
For Russia, which has traditionally linked information operations to broader geopolitical strategy, control over automated information systems would provide a strong lever—not only to impact public opinion today, but also to influence the AI chatbots that organize knowledge tomorrow.
4. A Growing Crisis of Credibility—and What Comes Next
The ISD analysis finishes with a clear warning: the Pravda network’s power is increasing not because it is clever, but because the digital ecosystem unintentionally supports it. Backlinks indicate trust. Engagement indicates significance. High volume promotes visibility. Even renowned publications with strict editorial standards have linked to Pravda content since it looks to be simply another news source among countless.
This legitimacy gap allows the Pravda network to affect narratives around Russia’s war in Ukraine, European policies, and global events. As mainstream media coverage of Ukraine continues to decline, experts fear that the network will take advantage of the gaps by filling them with Kremlin-aligned content. And these developments might create a possibility that the Russian viewpoint will become the dominant one available to AI systems, researchers, and even journalists looking for sources online.
For the time being, ISD has developed a browser extension that informs users when they visit a Pravda network site by changing the color of the page and adding a warning label. However, this is only a partial solution. The larger, structural issues remain unresolved, including how to prevent search engines from elevating propaganda, how to ensure that academic and journalistic institutions thoroughly vet sources, how to protect LLMs from absorbing manipulated content, and how to reduce the attraction for websites to link to high-volume network content in the first place.
The spread of the Pravda network highlights a bigger dilemma for the global information landscape. Propaganda of authoritarian regimes no longer spreads solely through persuasion or craft; it now spreads through infrastructure, scalability, automation, and unintentional amplification by respectable organizations. As long as algorithms reward volume and backlinks, operators that flood the system will continue to succeed.
The ISD analysis finally shows that the Pravda network is more than just a propaganda platform; it is an information supply chain. And unless governments, media organizations, and AI platforms respond swiftly, the network’s impact will increase, influencing not only what individuals read but also what the world’s most powerful AI systems learn.
