A concerning trend has emerged in the online information landscape, with hundreds of English-language websites, ranging from established news outlets to niche blogs, inadvertently amplifying pro-Kremlin propaganda. This revelation comes from a study conducted by the Institute for Strategic Dialogue (ISD), a London-based think tank. The study's findings highlight a disturbing reality: these websites, in their citations, treat the pro-Kremlin network as a credible source, thereby legitimizing its narratives and increasing its reach.
The network, known as the Pravda network, was identified by the French government last year as a disinformation operation. And here's where it gets controversial: by linking to articles from this network, even when disputing its credibility, these websites are increasing the chances of search engines and large language models (LLMs) encountering and promoting these pages.
Security experts have raised alarms about Russia's potential efforts to 'groom' chatbots like ChatGPT and Gemini with pro-Russia narratives. This process, known as 'LLM grooming', involves feeding these models large volumes of disinformation. The Pravda network, which has been active since 2014, has seen a significant surge in its output this year, with researchers reporting a daily publication rate of up to 23,000 articles in May, compared to approximately 6,000 in 2024.
The network's expansion is not limited to Russia; it now targets countries across Asia, Africa, and Europe, aiming for a global audience. Nina Jankowicz, a disinformation expert, warns that the network is targeting various languages and countries to establish a widespread presence.
What's driving this increase in activity? Some experts believe it's an attempt to influence the training datasets of AI models, which rely on vast amounts of data, including content scraped from the entire internet. Studies earlier this year revealed that popular chatbots sometimes repeated Russian disinformation in response to certain queries, suggesting, for instance, that the US was developing bioweapons in Ukraine or that the French were supplying mercenaries to Kyiv.
Researchers at the ISD argue that, regardless of whether LLMs have been poisoned, the Pravda network's high-volume strategy appears to be effective. Joseph Bodnar, a senior researcher at ISD, states, "More than any other Russia-aligned operation, the Pravda network is playing a numbers game. They've saturated the internet ecosystem to the point where they're getting in front of real people researching Russia-related issues."
The ISD's study found that 40% of the Pravda network's content picked up by mainstream websites was related to Russia's war in Ukraine. However, a significant portion of the content covered other topics, such as US domestic policy and Elon Musk's ventures. The Pravda articles have not only appeared on news websites but also on social media platforms.
Bodnar notes, "This happened across the board, affecting reputable and less reputable sources alike, spanning the ideological spectrum. It truly touched every corner of the web that we could find."
Jankowicz warns that the Pravda network's growing legitimacy could lead to a shift in media coverage, with the network potentially 'usurping' coverage on Ukraine as media outlets shift their focus elsewhere. She adds, "There's less news about Ukraine these days. If the Pravda network can fill that gap quickly, the Russian viewpoint will be the one that gets out there first and is cited in large language models."
This development raises important questions about the role of online platforms and the responsibility of content creators in the age of AI and disinformation. What are your thoughts on this issue? Feel free to share your opinions and engage in a discussion in the comments section below.
