The decision by Colombia’s National Electoral Council (CNE) to ban the publication of polls by Atlas Intel in Semana magazine has opened a new chapter in the debate over the transparency of political polling in Colombia during the final stretch of the presidential campaign.
The electoral authority believes there are indications of possible irregularities in the dissemination of opinion studies that, according to the agency, may not comply with the technical and legal requirements established for electoral polls.
The measure, adopted amid an increasingly polarized political climate, has sparked mixed reactions among political sectors, analysts, and media outlets. While some interpret the decision as a defense of electoral integrity and the information received by citizens, others warn about the risks of an administrative authority limiting the publication of opinion studies during a democratic process.
From positions close to the aforementioned media outlet, the CNE’s decision to open a preliminary investigation has been labeled “censorship,” in a case that was placed in the hands of the no-less controversial magistrate Alvaro Hernan Prada, who forwarded the report and observations to the polling firm.
Controversy in Colombia over alleged misuse of polls
The case revolves around Atlas Intel, an international data analysis and public opinion firm that in recent years has gained notoriety for its electoral measurements in various countries across Latin America and the United States. In Colombia, its polls began to be disseminated by Semana magazine during the current electoral cycle.
According to the CNE, some of the published measurements may contain inconsistencies related to the methodology used and the very nature of the studies disclosed. The electoral authority maintains that there are doubts as to whether certain exercises presented as polls actually correspond to opinion surveys conducted under the standards required by Colombian legislation.
The electoral body also noted that, under current regulations, every electoral poll must comply with specific requirements regarding technical identification, surveyed universe, sample size, margin of error, financing, and statistical procedures. The CNE’s concern is that some content presented as polls may in fact be digital exercises or open-participation measurements that would not have the same methodological rigor.
The decision was formalized in a resolution signed by magistrate Fabiola Marquez, who adopted part of the conclusions of a technical report prepared by the Technical and Oversight Commission on Public Opinion and Electoral Polls.
“To decree as an urgent precautionary measure the immediate suspension and/or temporary prohibition of the publication, dissemination, and divulgation of electoral polls by the polling firm Atlas Intel S.A.S. and the media outlet Semana magazine,” states the resolution issued by the CNE.
The resolution issued by the electoral authority ordered the suspension of the dissemination of new Atlas Intel studies in Semana while the corresponding verifications are carried out. The administrative process seeks to determine whether there were violations of the rules governing the publication of polls during electoral campaigns.
Aquí parte del AUTO
RADICADO CNE-E-DG-2026-014724
(19 de mayo de 2026) del CNE que según Semana, es una persecución.
– En el auto queda claro que no eran encuestas sino sondeos
– Que no era un error sino un patrón para posiblemente manipular al elector
– Además son tres… pic.twitter.com/GZO5xzRmcH— Alejo TORO (@AlejoToroAnt) May 19, 2026
Questions surrounding the so-called ‘surveys’
One of the central points of the debate is the difference between an electoral poll and other mechanisms for measuring opinion on the internet. Public opinion specialists explain that traditional polls require representative samples and verifiable methodologies that allow results to be extrapolated to the general population.
In contrast, so-called digital surveys are usually based on voluntary participation by users on virtual platforms, which can generate significant biases. For that reason, Colombian legislation establishes clear limits on how these types of exercises may be presented to the public.
The complaint that led to this decision was filed by Historic Pact representative Alejandro Toro over a poll published on April 9 in Semana magazine.
According to the congressman, the study contained inconsistencies related to the margin of error and the statistical method used to construct the sample. In the complaint filed before the CNE, Toro warned that “the poll reports a margin of error of ±2% with a 95% confidence level, without the sampling method used corresponding to a strict probabilistic design, which raises doubts about the statistical validity of the reported margin of error.”
Despite all this, the CNE has not definitively stated that Atlas Intel committed irregularities, but it does believe there are sufficient elements to open an administrative investigation and adopt preventive measures while the facts are clarified. The authority insists that its actions seek to protect citizens’ right to receive clear and verifiable electoral information.
The discussion has also highlighted the growing impact of digital platforms and new data analysis methodologies in modern political campaigns. In different countries, electoral authorities have had to update their criteria regarding technological tools that do not always fit within the traditional categories of polls and opinion studies.
Semana and Atlas Intel’s response
After the decision became public, Semana questioned the actions of the National Electoral Council and defended the legitimacy of the published measurements. The outlet maintained that the disseminated polls were produced by an internationally recognized firm and argued that the measure could be interpreted as a restriction on journalistic work and freedom of information.
Atlas Intel, for its part, has on previous occasions defended the accuracy of its statistical models and its methodology based on digital tools and massive data analysis. The company has highlighted accurate results in electoral processes in other countries, although its methods have also generated controversy among traditional polling experts.
So far, the discussion remains in the administrative arena, and there is no final decision regarding possible sanctions. The CNE process must determine whether there were indeed regulatory violations or whether the publications complied with current rules.
For his part, Atlas Intel CEO Andrei Roman added a comment defending his company’s work, asserting that “Paloma Valencia will not make it to the runoff,” a spot that his study projects will go to Abelardo De la Espriella who, according to the firm’s work, would contest the presidency against ruling-party candidate Ivan Cepeda in the June 21 runoff.
A sensitive issue during the election campaign
The debate takes on special relevance because of the political weight polls carry during presidential campaigns. In Colombia, as in other countries, polls influence public perception regarding candidates’ momentum or decline, affect strategic campaign decisions, and can alter funding dynamics and alliances.
Precisely because of that impact, electoral polls are subject to specific regulation. Colombian legislation requires transparency regarding who finances the studies, how they are conducted, and what their statistical limitations are. The objective is to prevent manipulation that could distort public opinion or create misleading perceptions about electoral preferences.
Electoral law experts point out that the challenge lies in balancing two fundamental principles: freedom of information and the need to protect the transparency of the democratic process. Excessive state intervention in polls could be interpreted as censorship, but the total absence of controls could also open the door to disinformation or political manipulation.
As the investigation moves forward, the Atlas Intel case has become a symbol of a broader discussion about the role of polls in the digital era and about the ability of authorities to regulate new forms of political measurement that evolve faster than existing rules.
Amid the presidential campaign, the controversy promises to continue fueling tensions between electoral bodies, media outlets, consulting firms, and political campaigns. The outcome of the process could set an important precedent regarding the legal limits of opinion studies in Colombia and the way electoral measurements will have to be presented in the coming years.