Against the backdrop of accelerating digital and information manipulation, computational propaganda poses a growing threat to information integrity and democratic discourse. Authoritarian actors and their proxies are increasingly leveraging algorithmic systems, search engines, and language models, among other technological solutions, to amplify misleading narratives at scale. This evolving landscape demands urgent scrutiny of the tactics, networks, and technologies fueling such influence operations.

The following report investigates the Bulgarian branch of the Pravda ecosystem — a sprawling network of websites that floods the Internet with pro-Kremlin content targeting over 50 countries, primarily in Europe. The analysis zooms in on Bulgarian Pravda’s web architecture and capacity to influence search engine algorithms and large language models (LLMs) through mass-produced, automated aggregator websites with carefully curated content and extensive search engine optimisation (SEO). Through a combination of OSINT collection, search ranking analysis, infrastructure mapping, and other digital forensic methods, the study corraborates the hypothesis that the Pravda ecosystem targets the algorithms, recommender systems and training data used in some of the most popular online services. In addition, Pravda acts as a web-based repository for the vast volume of pro-Kremlin content published on certain Telegram channels, themselves serving as automаtic aggregators of Russian government actors and Kremlin-controlled media. The findings underscore the urgent need for coordinated countermeasures, algorithmic transparency, and enhanced informаtion resilience across Europe and beyond.