BANNER: The logo of Russian internet group Yandex is pictured at the company’s headquarter in Moscow, Russia, October 4, 2018. (Source: REUTERS/Shamil Zhumatov)

Since the onset of Russia’s war on Ukraine, global technology firms have grappled with how to mitigate disinformation and respond to requests from the Russian government to censor information on their platforms. While much international attention has been paid to social media platforms, much less has been given to Yandex, the most popular search engine and news aggregator among Russian internet users. Yandex has promoted state media and narratives in its search results, and de-ranked and removed content critical of the Kremlin.

While Google remains the dominant search engine globally, Yandex is the leading search engine within Russia, the most popular search engine for news reading, and the most visited online resource in the country, with over 84 million monthly users. International technology firms leverage Yandex for their own platforms to provide a Russian-language search experience, such as with Mozilla’s Firefox browser, or generating non-news links for users in certain markets, such as for DuckDuckGo’s sites in Russia and Turkey. Both of these companies have recently severed or paused their relationships with Yandex.

Long before the current war, the Russian government put in place a series of laws giving state censor Roskomnadzor the legal mandate and technical capacity to curtail online information and discourse. In 2012, the Russian State Duma passed an amendment that empowered various government agencies to contribute to a “blacklist” of websites believed to be promoting content harmful to minors, for Roskomnadzor to block without a court order. In 2013, additional legislation allowed for the pre-trial blocking of “extremist content,” as well as content promoting “riots and illegal meetings, or any other “activities carried out in violation of the established order.” The broad wording of both pieces of legislation empowered Putin’s government to dictate which content is allowed on the internet and order internet service providers to block the websites hosting it without due process.

In 2017, the Duma passed a law on news aggregators requiring search services and their owners to check the accuracy of information displayed on their platforms, with a carve-out for content from state-registered media. The law also limits foreign ownership by requiring that the owner of a news aggregator is a Russian citizen or legal resident registered with the Russian government. By creating liability for platforms for hosting content from outlets that have not been vetted by the Kremlin, online service providers are deterred from showing news from non-licensed sources. In addition to other laws restricting journalistic freedom, this law targets the underlying system that distributes and determines what information is most readily available to internet users. To operate as widely as it does in Russia, Yandex complies with these domestic regulations, impacting its approach to content moderation and use of algorithmic ranking. Yandex dropped unofficial sources, such as blogs and foreign media, from its aggregator and began to prioritize state-approved outlets. In other words, the law has created a chilling effect.

Since Russia’s invasion of Ukraine, the crackdown on freedom of expression in the country has accelerated. The Duma passed an additional law introducing criminal liability for sharing “fake” content — i.e., any content the government deems to be “fake,” regardless of veracity — related to Russia’s armed forces, with a punishment of up to fifteen years in prison. In practice, the law is meant to halt critical coverage of the war, punish journalists and citizens speaking against it, and further insulate the Russian populace from accurate information.

To understand how Russian internet users are consuming information online and the impact Yandex policy and design has on that information, the DFRLab used the anonymous web browser Tor to simulate searches on Yandex with a Russian IP address, so the results returned from those searches would be the same as those seen by someone living in the country. When compared to other search engines, Yandex consistently showed higher rankings of common Russian disinformation narratives, de-ranking of content and media indicating opposition to the war, and a complete lack of access to sites labeled as “restricted” by the Kremlin.

According to Yandex, “The algorithm determines what goes into the top Yandex.news page by prioritizing the licensed media with the most citations, the fastest speed of news delivery, and the greatest audience engagement.” If you run certain searches on Yandex — for example, “война в украине” (“war in Ukraine”) — from a Russian IP, it warns users that “some links are missing in the search results due to the requirements of the legislation of the Russian Federation” and offers links to a support page citing the 2017 law news aggregator law. This means that Yandex excludes from its results any sites that Russia has deemed “restricted” by Roskomnadzor. From a US IP address, these sites do appear in Yandex search results, but they are indexed lower than they are on Google, for example.

Additionally, Yandex has added a content label to search results about the war in Ukraine, telling users “Some material on the internet may contain inaccurate information. Please be attentive.”

Russian users searching Yandex will find that links to Western and domestic outlets critical of the regime are unavailable, including the BBC, Meduza (a Russian- and English-language independent news outlet), and Radio Free Europe.

Yandex is joined in this targeted moderation by VKontakte, one of Russia’s largest social media platforms, which blocks the same outlets as Yandex.

Similarly, searches on Yandex for social media platforms banned by the Russian government return zero results for users with a Russian IP address. In contrast, when conducted via a US IP address, Yandex search results still include social platforms, indicating that the ban is not platform-wide but isolated to just those accessing the search from a Russian IP address.

Yandex’s restriction of certain sites is not particularly surprising, as Russia has cut off internet access to these sites on its own. Nonetheless, when Russian users search Yandex for information, the vast majority of opposing or unbiased results have disappeared, leaving them with a stream of content pushing the same pro-Kremlin narratives.

Google is the second most used search engine in Russia after Yandex. Although Google has not complied with censorship laws to the same extent as Yandex, it recently told Russian translators not to refer to the invasion of Ukraine as a “war” to avoid legal ramifications. On March 23, Roskomnadzor blocked Google News, claiming it was providing access to inauthentic material on the “special operation” in Ukraine. As all research for this piece was conducted before the Google News ban, many of the comparisons made here would now be difficult to replicate on Google from a Russian IP address.

Blocking Google News is not the same as blocking Google’s search engine; Google News is a news aggregator, which means it collects and displays articles from many different outlets. It can be viewed through an app, website, or through its search engine results. Russia’s ban will prevent in-country users from seeing Google News results when they conduct a search.

From March 16 to 17, the DFRLab ran a variety of terms through Yandex’s search engine, including “Ukraine” (“Украина”), “biolabs” (“биолаборатории”), “protests” (“протесты”), and “no war” (“нет войне”). We compared findings to Google search engine results retrieved using a spoofed Russian IP address. Most notable was the disparity in results shown on Yandex News versus Google News.

Across all news searches, Yandex showed a clear tendency to amplify — especially within the top ten to twenty results — propagandist and pro-Russian sites such as iz.ru, Ukraina.ru, and life.ru, along with Russian state media. Comparatively, Google returned a mix of Russian state media, Western or anti-war content, and fact-based reporting. For example, when searching “Ukraine” on Yandex, the top ten results contained exclusively Russian media sites pushing a pro-Russian slant, including two stories amplifying the biolab narrative, which is built around the debunked claim that there are US-run biolabs in Ukraine developing bioweapons to target Russia or Russian-speaking populations.

Similar results occurred when searching for the word “protests” on Yandex. Within the top ten results, only one source appeared to give a fact-based report of the ongoing anti-war protests in Russia, while the other nine articles reported on either law enforcement issues (e.g., the number of protesters arrested) or protests occurring elsewhere in the world. This was a stark contrast to the first page of Google results, in which every link discussed Russian anti-war protests and the detainment of protesters.

The DFRLab also analyzed differences in the image tabs of Yandex and Google based on the same search terms. When examining image results for the term “Ukraine,” we noticed limited presence of war-related images in Yandex search results versus those displayed on Google. Yandex images, displayed below, showed mostly Ukrainian flags and landscape photos, along with an image of a swastika superimposed over blue and yellow, Ukraine’s national colors. Google shows a wide variety of images, including scenes from the ongoing war.

The sources of these photos also provide insight into Yandex’s photo indexing algorithm. Yandex seems to be giving more weight to images from older articles and sites; of the first fifteen image results, none were published in 2022, and none mention the war. Most of the images come from 2020–2021 articles or from Ukraine tourism and fact sites. Although many of Google’s first fifteen image results come from Russian news outlets, these articles were mainly published between December 2021 to March 2022 and included Western outlets such as the BBC and Radio Free Europe.

After examining the results of general keyword searches about Ukraine and anti-war sentiment, we wanted to see what the information landscape would look like for a more specific incident. To test this, we searched for “Marina Ovsyannikova” (“Марина Овсянникова”). Ovsyannikova was an editor for Russian state-owned Channel One who crashed its nightly live news broadcast with a sign reading “No War,” in direct violation of the recently passed law making it illegal to say those words in relation to the Ukraine invasion.

When searching her name on Yandex, no images or video of Ovsyannikova appear. Even when reverse-image searching a commonly used screenshot from the video of Ovsyannikova crashing the broadcast, Yandex does not bring up a single instance of the same image. This is highly unusual, as Yandex is known among the open-source research community for its impressive reverse image search capabilities. This search was conducted from a US IP address, which suggests that Yandex’s censorship of some content is not exclusive to Russia.

In a Facebook post addressed to his former colleagues, Lev Gershenzon, who previously served as the head of Yandex’s news division, wrote, “On the main page of Yandex at least 30 million Russian users see that there is no war, there are no thousands of dead Russian soldiers, there are no dozens of civilians killed under Russian bombing, there are no dozens of prisoners, there is no huge destruction in various Ukrainian cities.” Our findings give credence to Gershenzon’s words and demonstrate how Russia’s legislative and regulatory frameworks have made Yandex a tool in Russia’s war against Ukraine and domestic opposition to the war.

Jacqueline Malaret and Ingrid Dickinson, “Yandex suppresses Ukraine war information for Russian internet users,” Digital Forensic Research Lab (DFRLab), March 31, 2022, https://medium.com/dfrlab/yandex-suppresses-ukraine-war-information-for-russian-internet-users-caadfed968b5.