For some time now, products have appeared in the ecosystem of think tanks and “observatories” funded to talk about Cuba that attempt to dress old political campaigns in the guise of technical neutrality.
The latest example is an investigation by the 4Métrica Foundation titled “Computational Analysis of Discourse on Changes in Perception of Cuban Medical Missions on the Social Network X,” presented as if it were the definitive proof that the Cuban state manipulates world public opinion about its medical cooperation.
On the laboratory’s own webpage, the study is promoted as follows: “If you want to know how we exposed the digital guerrilla operations led by the Cuban regime, we invite you to read the investigation.” This hardly sounds like the language of a neutral center approaching data without prejudice, but rather that of an organization that has already decided the verdict and is now looking for the flashiest method to justify it.
From 1,644 tweets to “global public opinion”
The core of the work is a set of 1,644 messages on X (formerly Twitter), in Spanish, supposedly collected between 2015 and 2025 with a single search query: “misiones médicas” OR “misiones médicas cubanas.”
These messages are divided into two groups: 481 tweets before 2020 and 1,163 after. That is, 13.7 messages per month over ten years. Based on this minimal sample, the “investigation” claims to be measuring “public opinion” about the missions and, even more, demonstrating a “reversal” in international perception caused by a Cuban digital operation.
“To examine the dynamics of public opinion (sic), a mixed-method methodological approach was designed. A corpus of 1,644 tweets was extracted from X’s API using the query = (misiones médicas OR misiones médicas cubanas), divided into two subsets: 481 for the period 2015-2019 and 1,163 for 2020 onward,” says 4Métrica.
The problem is obvious.
First, X is not “global public opinion,” but a platform with very strong biases of class, language, geography, and ideology.
Second, reducing the global conversation about Cuban medical cooperation to 1,644 messages in Spanish over ten years is simply ridiculous when compared to the actual magnitude of the phenomenon: tens of thousands of professionals, dozens of countries, coverage in traditional media, statements from governments, parliaments, international organizations, campaigns for the Nobel Prize for the Henry Reeve Brigade, and so on.
Third, the very search is so narrow that it excludes a large part of the conversation from the outset: “Cuban doctors,” “medical brigades,” “Henry Reeve Brigade,” “Cuban doctors in Italy/Mexico/Brazil,” “Cuban medical cooperation,” and so many other phrases used by media, governments, and citizens to talk about the same topic.
The result is a small, biased, and unrepresentative corpus that is nevertheless presented as if it were a universal thermometer.
Algorithms without context: the new “objective” excuse
To give a modern touch to the narrative, the study resorts to buzzwords: “network analysis,” “topic models,” “generative AI.” Interactions on X are mapped, and as is logical in any social network, dense clusters appear around official Cuban accounts, Island media, and solidarity networks. This is immediately described by the "research" as “computational propaganda” and “digital guerrillas” serving the “regime.”
What is never done is applying the same criteria to other actors on the board. In the 4Métrica repository itself, we find annual reports on Cuba, policy papers, documents from the so-called “Social Rights Observatory,” and other platforms aligned with the “regime change” agenda.
In other words: there is a constellation of organizations, NGOs, and media that systematically work against the Cuban state, often with funding from U.S. federal agencies, political coordination, and articulated presence on social networks. But this is never mentioned in the analysis as possible “computational propaganda.”
The double standard is evident: when the Cuban state communicates, it is manipulation; when Western governments, private foundations, or groups funded by the U.S. government do it, it is “civil society.” When thousands of people share content favorable to Cuba, they are “digital troops”; when campaigns against the country are organized, they are “freedom activists.”
Even more serious: the artificial intelligence model used to classify tweets as "favorable," "neutral," or "critical" is a black box. It does not specify what system is employed, how it was trained, what its accuracy is, or provide a single validation metric. Nevertheless, those numbers are averaged and turned into graphs that supposedly prove a "critical" before and a "positive" after in perception.
That is not science; it is marketing with graphics.
The elephant in the room: the pandemic and international recognition
If one looks at the real world—not just a corner of X—the change in tone in many narratives about Cuban medical missions has a much simpler explanation than a digital conspiracy: the COVID-19 pandemic.
In 2020, while some governments were fighting over masks at airports and others were competing for vaccine patents, Cuban medical brigades were arriving in Lombardy, Piedmont, Andorra, and other regions hit by the virus. In Italy, a central country in the European imagination, images of doctors and nurses from the Henry Reeve Brigade were received with gratitude by local authorities and the population, generating a wave of sympathy that is still remembered today.
It was not an isolated episode. Since its creation in 2005, the Henry Reeve contingent has responded to emergencies and disasters in more than twenty countries, has attended to millions of people, and received the Dr. Lee Jong-wook Public Health Award from the World Health Organization and PAHO in 2017, precisely for its solidarity work in epidemics and disasters. Various organizations, movements, and personalities have supported the contingent's nomination for the Nobel Peace Prize in recent years.
Is it really necessary to have a lot of "computational propaganda" to explain that, in that context, the number of messages speaking positively about Cuban medical cooperation has increased? Or is it rather that what bothers certain observers is that the concrete practice of solidarity amid a global pandemic contradicts the caricature of Cuba as a "slave state" that they are trying to impose?
What the 4Métrica study does is exactly that: it takes for granted, as "documented reality," the narrative that missions are "modern slavery," relying almost exclusively on hostile organizations like Prisoners Defenders (an appendage of UNPACU, funded by federal agencies of the U.S. government) and certain politicized reports; while systematically ignoring the voices of recipient governments, international health organizations, and social movements that recognize the solidarity and voluntary nature of that cooperation.
From "public opinion" to the psychological warfare laboratory
There is a political aspect underlying this that cannot be overlooked. 4Métrica is not an isolated actor. In its own presentation, it defines itself as a publishing label "in defense of freedom and democracy in Latin America" and hosts products like Cuba Report 2022, Cuba Report 2023, or documents drafted with organizations from Florida that openly work for regime change on the Island.
In other words, we are facing yet another node in a network of think tanks, observatories, and platforms that divide tasks: some produce "reports" on human rights, others "surveys" on informal economy, others "computational analyses" on social networks. They all share the same premise: Cuba is a "dictatorship," its social policies are "facades," its doctors are "slaves," and any narrative that does not stem from this must be unmasked as propaganda.
In that logic, "public opinion" is not something to be understood in its complexity but rather a battlefield that needs to be dominated. If reality, healthcare statistics, or the concrete gratitude of peoples and governments do not fit the script, they resort to pocket statistics, context-free algorithms, and bombastic words like "digital guerrillas" or "influence operations" to delegitimize any differing account.
Behind it all is the same old story: disinformation and manipulation as a weapon of war.
