The Information Laundromat
In May 2024, ASD, along with our partners at the Institute for Strategic Dialogue and the University of Amsterdam, published a report documenting roughly 400 domains–ranging from mirror sites and content aggregators to faux local news outlets and sites ostensibly focused on spirituality and men’s interests—that republished articles that were identical or nearly identical to those that originated on Russian state media outlet RT.com. That report highlighted the urgent need for an ongoing monitoring system to detect websites laundering content from Russian state media outlets—particularly those targeting audiences in the United States.
To that end, we have paired two of ASD’s tools: the Hamilton 2.0 dashboard, which tracks, among other things, articles published by Russian state-controlled media outlets, and the Information Laundromat, a tool that can help detect duplicate or near-duplicate content on the open web.
This dashboard aims, in part, to help secure the 2024 US elections by providing near real-time insights into how Russian state-sponsored propaganda infiltrates and influences our digital information environment. By monitoring where content from RT and RT en Español is being republished or amplified across the open web, this system can identify Russian content laundering sites before they reach a wider audience in the United States. This will allow stakeholders to implement countermeasures, including but not limited to flagging content for fact-checkers or raising public awareness about websites acting as fronts for Russian propaganda.
In this critical election, every narrative and piece of information can impact voter perceptions and behavior. This tool is a frontline defense against foreign information manipulation undermining the democratic process.
Each day, we will run recently published articles from RT and RT en Español through the Information Laundromat to identify where that content has been reposted or cited. This tracking is visualized in a dynamic dashboard that presents data in two main ways:
The detailed explorer section provides a granular look at individual instances of RT content laundering, allowing users to investigate specific articles and where those articles were seemingly reposted. This continuous monitoring process ensures that the spread of Russian messaging, and any potential disinformation narratives, can be tracked in real-time, providing crucial insights into how Russian state media content proliferates across the web.
Similar to the methodology employed in the “The Russian Propaganda Nesting Doll” report, we employ thresholds to minimize potential spurious matches. For a domain and the accompanying results to be included on the dashboard, it must have matched a queried RT article at least 10 times with any average match score, six to 10 times with an average match score > 60%, two to five times with an average match score > 75%, or once with an average match score >90%. (Read more about match scores here). Each individual result is kept only if exceeds a 60% match score with the source query (without regard to the domain).
Limitations of Automated Analysis: While The Information Laundromat uses sophisticated algorithms to identify content similarities and connections, it cannot fully account for the context or intent behind the dissemination of information. Users should consider the broader context in which content is shared and be cautious in drawing conclusions based solely on automated matches.
Potential for Errors: The tool may produce false positives, where content appears to be laundered but is not, or false negatives, where laundered content is missed. Although we apply stringent filtering to minimize these occurrences, users should critically assess the results and, where possible, corroborate findings with additional sources.
No Assumption of Malice: The presence of content from state-sponsored media on a given website does not necessarily imply coordination or intent to deceive. Some websites may unknowingly republish content or do so for reasons unrelated to disinformation. Users should avoid assuming malintent without further evidence.
Difficulty in Proving Coordination: Identifying that a website has republished state-sponsored content is not the same as proving that there is a deliberate campaign to launder information. Attribution of intent or coordination requires more in-depth investigation and should not be inferred solely from the data provided by The Information Laundromat.
Adaptability of Adversaries: State-sponsored actors and other entities involved in information laundering are continuously evolving their tactics. The information and connections identified by The Information Laundromat reflect a snapshot in time and may not capture the full scope of ongoing disinformation campaigns.
Varying Interpretations: The significance and impact of laundered content may vary depending on regional and cultural contexts. Users should be mindful of these differences when interpreting results, especially when analyzing content that crosses linguistic or national boundaries.
Search Engine Variability: The tool relies on results from various search engines, which may differ in their indexing, ranking, and censorship practices. This variability can affect the results and should be taken into account when interpreting the data.
Not a Conclusive Source: The Information Laundromat is intended as a research aid and not as a definitive source of truth. It is best used as a starting point for further investigation rather than a final determination of the origins and spread of content.