The Algorithmic Homeland: AI, Streaming, and Affective Mobilization in Colombia’s 2026 Far-Right Presidential Digital Campaign

Abstract

This essay draws on a delimited corpus of publicly circulating materials associated with Abelardo de la Espriella’s digital campaign in the 2026 Colombian presidential election to examine a form of algorithmic–affective campaigning. Through a qualitative, discursive, and multimodal approach, it examines a core analytical corpus comprising the complete stream with Westcol and six digital posts, supplemented by materials from a broader documentary mapping and journalistic coverage used to contextualize provenance, circulation, and controversies. The central argument holds that the campaign used streaming and generative AI as technological mediations to organize and intensify patriotic and religious language of threat, protection, certainty, belonging, and authenticity. Through the streamed interview with Westcol, a Colombian creator known for gaming content and a large online presence, the campaign staged performative authenticity and proximity through digital codes associated with youth cultures; its use of generative AI contributed to the construction of visual narratives of threat, order, and national salvation. The article develops the notion of the algorithmic homeland and defines it as an imaginary of the nation represented as a threatened, morally ordered community in need of protection, configured through platform infrastructures, media formats, and affective languages. This concept makes it possible to interpret a form of electoral legitimacy construction in which technological mediation and affective languages jointly participate in the production of visibility, proximity, and authority within platform democracy.

1. Introduction: From the Digital Campaign to the Algorithmic Homeland

The 2026 Colombian presidential election unfolded amid a polarized dispute over the political succession to Gustavo Petro and culminated in a closely contested runoff between left-wing senator Iván Cepeda and far-right candidate Abelardo de la Espriella, who won by a narrow margin (Suárez, 2026). A lawyer and entrepreneur with no previous experience in public office, De la Espriella presented himself as an outsider to professional politics through the Defensores de la Patria movement (Acosta, 2026). From the registration of his signature-gathering committee in July 2025 to the runoff on 21 June 2026, his campaign combined promises of hardline security, accelerated economic growth, and a reduced state with appeals to homeland, family, religion, and moral order (Villa Román, 2026). These themes circulated through the spectacular persona of “El Tigre,” social media content, influencer collaborations, streamed conversations, and AI-generated audiovisual pieces (Hernández Bonilla, 2026; Osorio, 2026a). This configuration situates the case within debates on contemporary Latin American political communication by bringing together digital temporalities, connected publics, synthetic images, national symbols, and the languages of entertainment in the production of proximity, certainty, and public authority.

Contemporary political communication is undergoing a reorganization of its mediations. The public square, the televised debate, and the journalistic interview continue to form part of the democratic ecosystem, although they now coexist with spaces in which visibility is produced through live streams, short videos, viral clips, and communities connected in real time. The logic of networked media helps to explain this shift, since platforms modify the ways in which political content is produced, distributed, and appropriated (Klinger & Svensson, 2015). This transformation is also related to a broader dynamic of hybrid political communication in which traditional actors, platforms, publics, and digital technologies intervene in the construction of the electoral event (Chadwick & Stromer-Galley, 2016). The campaign thus appears as a communicative architecture in which the candidate’s message circulates alongside comments, audiovisual excerpts, user appropriations, and interpretive disputes that extend its public life.

The specificity of each platform occupies a decisive place in this process. Digital architectures shape campaign strategies by organizing differentiated modes of interaction, visibility, and user behavior (Bossetta, 2018). These technical affordances and conventions of use favor different forms of political presentation and circulation: TikTok tends toward short-form audiovisual content, Instagram toward visual forms of self-presentation and identification, and X toward rapidly circulating public commentary and controversy; YouTube supports the hosting and redistribution of longer-form audiovisual material, and streaming incorporates synchronous interaction between creators and audiences. Within this composition of media environments, the candidate circulates as a public figure, narrative, and affective presence configured differently according to format. Politics thus enters an environment in which each platform establishes particular conditions for attracting attention and producing different forms of proximity, controversy, and spectacularization.

This article proposes the category of algorithmic–affective campaigning to interpret De la Espriella’s digital strategy. The algorithmic dimension refers to conditions of visibility and circulation configured by platforms through recommendation systems, interaction metrics, and the temporalities specific to each digital environment. The affective dimension concerns the languages through which the campaign organizes messages of threat, protection, belonging, certainty, and authenticity. In their articulation, algorithmic–affective campaigning generates a political experience in which messages acquire intensity through their circulation in formats adapted to platform logics and through emotional repertoires capable of producing proximity and identification. This category makes it possible to understand how technological mediation participates in the amplification and reorganization of affective languages within electoral communication.

The term algorithmic homeland also contains a deliberately ironic displacement of the “Patria Milagro” (“Miracle Homeland”), the expression through which De la Espriella articulated his project of national reconstruction during the campaign (Defensores de la Patria, 2026). The providential dimension of this formulation became particularly visible on 13 June 2026, when Christian singer Alex Campos joined the candidate in a conversation circulated under the title “God’s Promise for Colombia: The Miracle Homeland” (Bustos Granados, 2026). During the encounter, the electoral contest was framed through a language of divine purpose, God-guided leadership, and spiritual battle, while the Miracle Homeland was presented as a promise of greatness, prosperity, and peace for Colombia (DE LA ESPRIELLA STYLE, 2026). The shift from miracle to algorithm operates here as a conceptual and critical gesture that redirects attention from the providential register of the electoral promise toward the sociotechnical mediations through which that promise acquired visibility, intensity, and circulation. The algorithmic homeland thus designates the way in which a national imaginary of threat, protection, and restoration was staged and redistributed through platforms, recommendation systems, streaming, influencer networks, and synthetic visualities, while maintaining the interpretive boundary whereby this study does not attribute the results of the election to having been determined by technological factors alone.

The force of this campaign modality intensifies in contexts marked by social uncertainty, institutional distrust, and exhaustion with the conventional language of politics. Within these contexts, affects participate in the ways that events acquire public meaning and particular narratives connect individual experiences with collective forms of identification. The notion of affective publics helps to understand this circulation as a narrative dynamic through which feelings, events, and practices of participation become interconnected within digital environments (Papacharissi, 2016). From this perspective, security, homeland, family, religion, and moral order acquire force when they translate dispersed concerns into emotionally recognizable narratives of threat, protection, belonging, and certainty. The persistent relationship between digital media and political participation further underscores the importance of examining these forms of engagement within contemporary communication ecosystems (Boulianne, 2020).

This relationship between affects and digital circulation has particularly relevant precedents in Latin America. In Colombia, research on the 2021 National Strike showed how interactions between political actors and Facebook users configured forms of affective attunement that articulated individual and collective feelings within public discourse (Torralba et al., 2023). In Argentina, Javier Milei’s presidential campaign deployed an emotional strategy across X, Instagram, and TikTok based on resentment and positive emotions associated with the construction of a political “we” (Montero, 2024). A comparative study of electoral campaigns in Brazil, Colombia, Chile, and Peru also found that divisive issues and users’ emotional reactions participate in dynamics of visibility and polarization on Facebook (Fenoll et al., 2024). De la Espriella’s campaign forms part of this regional landscape while presenting a particular configuration through its combination of these affective repertoires with streaming, influencer mediation, and content produced through generative artificial intelligence.

In this scenario, streaming introduces strategic mediation. The figure of the streamer translates politics into a lexicon of conversation, reaction, proximity, and entertainment. His authority derives from the accumulated familiarity with communities that watch, comment on, and recirculate the content. Performative authenticity, studied in Enli (2017) in campaigns shaped by social media, operates as a resource for projecting proximity while distancing the candidate from the institutional codes of professional politics. In the case under study, this authenticity is articulated with the temporality of the live broadcast and with the streamer’s capacity to turn the candidate’s exposure into a shared experience. Streaming thus provides a scene of legitimation in which the attention accumulated around the digital creator creates new conditions of exposure, proximity, and recognition for the political figure.

Generative artificial intelligence introduces an additional technological mediation within this configuration. Its relevance lies in the possibility of producing images, sounds, voices, scenes, and montages that accelerate the spectacularization of political narrative. AI makes it possible to turn the electoral message into a visual epic, dramatized threat, promise of salvation, or caricatured representation of the adversary. This plasticity expands the boundaries of propaganda by introducing zones of ambiguity between fiction, satire, montage, persuasion, and disinformation. Synthetic political videos can affect public trust by increasing uncertainty about the authenticity of content (Vaccari & Chadwick, 2020). At a broader level, Jungherr (2023) argues that the relationship between artificial intelligence and democracy requires examining AI’s effects on information, public communication, participation, and political power.

The infrastructural dimension of this process is also relevant. Digital platforms participate in the production of political communication through their recommendation systems, interaction metrics, interface designs, and commercial infrastructures that influence how content becomes visible and circulates. van Dijck and Poell (2013) propose understanding social media through a logic that incorporates programmability, popularity, connectivity, and datafication. In the electoral field, technology firms, platform services, consultants, and digital teams form part of a broader environment through which campaigns produce visibility and interact with their publics (Kreiss & McGregor, 2018). In this article, these infrastructural conditions operate as the sociotechnical context of circulation rather than as evidence that De la Espriella’s campaign employed specific forms of behavioral targeting. The digital campaign can therefore be understood as a communicative configuration in which human actors produce, interpret, and recirculate content within platform architectures that rank, recommend, measure, and redistribute visibility.

Methodological Approach, Corpus, and Interpretive Procedure

This essay adopts a qualitative, interpretive case study approach. Its scope is limited to a bounded corpus and does not claim to provide a comprehensive mapping of the campaign or a causal assessment of its electoral effects. The inquiry examines how streaming and generative artificial intelligence operated as technological mediations within the digital strategy, while influencers and connected digital activists participated in the circulation and reformulation of content associated with the presidential candidate. The analysis also considers how patriotic, religious, and security repertoires organized affective languages of threat, protection, certainty, belonging, and authenticity. Abelardo de la Espriella’s campaign was selected because this relationship between technological mediations, circulation actors, and affective repertoires acquired a particularly visible expression within its digital strategy. This configuration is interpreted through two related concepts. Algorithmic–affective campaigning refers to the vocabulary of production, mediation, and circulation that links platform resources to emotional appeals. The algorithmic homeland refers to an imaginary of the nation as a threatened, morally ordered community in need of protection; this imaginary is generated and circulated through that vocabulary.

An initial documentary mapping identified more than twenty traceable digital materials circulating during the observation period. These materials originated from the candidate’s accounts, explicit collaborations, connected digital activists, and other forms of supportive circulation. From this broader documentary pool, a core analytical corpus was delimited through a purposive selection of seven primary units corresponding to an observation period extending from 12 March to 21 June 2026. It comprises the complete stream with Westcol, divided into seven thematic sequences, and six posts associated with the campaign’s circulation and published through the candidate’s accounts, explicit collaborations, or accounts supporting his candidacy. Materials from the broader mapping that fall outside these seven units are used selectively as supplementary documentary evidence to identify the recurrence, extension, and supportive circulation of the repertoires observed. Intensive multimodal interpretation is concentrated on the core analytical corpus, while these Supplementary Materials situate its patterns within a broader field of digital circulation. The provenance and documented relationship of each item to the official campaign are specified in the analysis. Materials were selected when their source could be traced, when they incorporated one of the mediations central to the essay, when they explicitly mobilized one or more of the affective languages under study, and when they articulated representations of the nation, authority, or belonging. The anthem “El tigre de la Patria,” published in 2025, is used solely as an antecedent to the tiger symbolism and remains outside the main corpus. Journalistic coverage provides contextual information about dates, circulation, and controversies but does not substitute for the close reading of digital materials.

The units of analysis were defined according to the format of each item. For the Westcol stream, the inquiry examined sequences delimited by changes in topic or setting; for digital posts, it considered each item as a whole, including accompanying text and audiovisual components. The multimodal reading attended to the interplay between verbal statements, visual framing, editing, music, gestures, patriotic and religious symbols, and platform-specific forms of participation. The procedure first identified the provenance, publication context, and relationship of each item to the campaign. It then examined the interaction between multimodal resources and affective framings. Finally, it interpreted how each item presented the candidate, his adversaries, and the national community and how these relations contributed to the essay’s two central categories. The claims are restricted to meanings and sociotechnical configurations observable in publicly circulating materials. Accordingly, they do not measure persuasion, establish the demographic composition of audiences, or attribute voting behavior to algorithms or particular campaign devices. Metrics, comments, and testimonies are used solely to describe visible forms of circulation and reception. Within these interpretive limits, the stream with Westcol provides a particularly productive entry point into the corpus because it contains, within a single mediated event, the relationship between platform format, the candidate’s performance, audience interaction, and the production of affective languages.

2. Streaming and Affective Campaign Vocabularies

The interview with the streamer Westcol contained, in a condensed form, one of the most visible dimensions of Abelardo de la Espriella’s digital strategy. Luis Fernando Villa Álvarez, known as Westcol, is a Colombian creator whose career trajectory began with gaming content and later expanded into conversational streaming and live entertainment, making him one of the most visible figures in this industry in Colombia and Latin America. By 2026, his accounts had also begun to incorporate highly visible political conversations, including an earlier interview with then-President Gustavo Petro (Ríos, 2026). De la Espriella appeared on Westcol’s Kick channel on 30 May 2026, on the eve of the first round of the presidential election and at the close of his campaign, on a stream that attracted more than 250,000 concurrent viewers and received coverage from national and international media (Hernández Bonilla, 2026). The choice of this timing and this platform made the appearance an observable communicative decision: the candidate moved the closing of his campaign into an environment already shaped by the conventions of live entertainment and by the sustained relationship between the creator and his community.

The stream also displayed a prepared staging that connected the candidate’s political identity with the streamer’s visual messaging. Its opening included an AI-generated sequence in which Westcol and De la Espriella appeared petting tigers in front of the Casa de Nariño, linking the host to the “El Tigre” symbolism used throughout the campaign (Osorio, 2026a; Hernández Bonilla, 2026). This incorporation makes the campaign’s strategic agency more visible: the appearance went beyond placing the candidate in an interview and turned the streamer’s platform into a scene compatible with the campaign’s own visual and affective repertoires. The conversation then unfolded through humor, reactions, questions, personal exposure, and live interaction, resources that shifted electoral presentation toward a form of proximity constructed within the format.

Performative authenticity constitutes one of the core elements of this mediation. Enli (2017) has shown that social media foster forms of political authenticity constructed through proximity, personal exposure, and distance from institutional codes. The Westcol stream makes this operation observable in a specific sequence. Between approximately 11:58 and 15:50, De la Espriella displays his Bible, recounts his religious conversion, and invokes the story of David and Goliath to interpret his own position within the political contest. The religious object becomes physically present within the streaming space, while the account is presented as a biographical experience, allowing the appeal to faith to take the form of a personal revelation shared before the audience. Streaming intensifies this operation because it places the candidate within a scene in which each response appears to unfold under the pressure of the live broadcast. The proximity generated by the format thus turns a religious reference into a scene of self-exposure: belief is presented through voice, gesture, the displayed object, and interaction with the host.

The camera, the chat, the silences, the interruptions, and the streamer’s reactions produce a sense of public testing. The candidate appears to be evaluated by his capacity to inhabit an open conversation, respond to uncomfortable questions, sustain the rhythm of the format, and project emotional control before an audience watching in real time. Within the religious sequence, a particular dimension is added to the scenario: the David-and-Goliath narrative presents vulnerability and strength within the same portrayal, positioning the candidate as a figure confronting a greater force through moral conviction and transcendent support. Religion operates here as an affective language of certainty, while streaming provides conditions of proximity and exposure through which religious conviction is presented as a personal experience before the audience. This scene brings together strategic preparation and an appearance of availability, making spontaneity, improvisation, and immediate response part of the performance of authenticity.

The same stream makes a second affective operation observable when the conversation shifts from personal exposure toward security. Between approximately 6:22 and 6:35, De la Espriella displays photographs of armed organizations and incorporates them into a representation of conflict organized around identifiable enemies. Later, between approximately 33:52 and 36:05, he characterizes insecurity as a “pandemic” and connects this perception of crisis with the defense of the family and the promise of authority exercised with an “iron hand.” The continuity between these sequences transforms a broad social problem into a more immediate narrative structure: threat acquires recognizable faces, danger is presented as a widespread experience, and the political response becomes concentrated in a figure capable of restoring order. Streaming contributes to this operation through the candidate’s extended exposure, allowing him to display images, develop the narrative, modulate his tone, and react to the interlocutor within the same communicative scene.

Fear operates here as an affective language that directs the perception of conflict and prepares protection as a political promise. The candidate’s authority acquires meaning within this relationship: as threat becomes increasingly visible, the representation of a leader prepared to confront it gains affective force. This strategy also connects with a repertoire present at other moments of the campaign, when criminal organizations were described through a language of confrontation and declared targets of a future security policy (Semana, 2026). Within the stream, this messaging also acquires domestic and personal dimensions by connecting security, family, and the capacity to protect. The sequence therefore contributes to one component of the algorithmic homeland: a community imagined through vulnerability, whose affective cohesion is constructed around the need for defense and the expectation that there is an authority figure who can provide it with certainty.

The live broadcast also reorganizes the temporality of the campaign. Networks operate through brief cycles of recall in which an image, phrase, or intervention is circulated intensely among the public for a limited period and is then displaced by new controversies. In this case, the specificity of streaming lies in combining the synchronous attention generated during the broadcast with the subsequent fragmentation of the event into clips, phrases, reactions, and posts that extend its circulation. De la Espriella’s appearance close to the election therefore acquired a double temporality: concentrated attention during the live event and the subsequent recirculation of easily shareable fragments. The logic of connective action helps explain this extension of the event through personalized and shareable expressions (Bennett & Segerberg, 2012). Comments, excerpts, and user appropriations prolong the public life of the content and facilitate its circulation through users’ own networks.

The generational dimension of this mediation lies in the cultural codes that streaming brings into political conversation. Westcol’s career trajectory through gaming and live entertainment places the presidential candidate within an environment shaped by simultaneity, informality, reaction, and visible audience participation, codes closely associated with youth digital cultures. Within this space, the streamer retains the capacity to set the rhythm, introduce topics, and translate questions for his community, while the candidate’s authority coexists with the host’s convening power. On the stream, therefore, a hybrid scene is constructed in which interview, entertainment, and real-time reaction produce a form of proximity distinct from the institutional political interview.

3. Generative AI, Patriotism, and the Production of Certainty

Generative artificial intelligence occupied a specific place within the strategy by expanding the visual possibilities through which antagonism, threat, and the promise of restoration could be represented. One AI-generated piece makes this operation particularly visible: De la Espriella appears transformed into a gigantic tiger that destroys a train associated with “Los de Siempre” (“the usual ones”), the expression through which his campaign identified traditional political actors. The scene turns an abstract political opposition into an immediate visual narrative. The adversary acquires a recognizable material form, while the candidate’s figure expands into a force capable of destroying it (Osorio, 2026c). Images, voices, montages, epic sequences, and highly emotionally charged compositions make it possible to present an interpretation of the country in forms designed for rapid circulation. In this case, AI transforms the antagonism between “los nunca” (“those who never had a turn”) and “los de siempre” into an audiovisual fantasy of confrontation and victory.

The specificity of generative mediation lies precisely in this capacity to produce scenes that exceed the possibilities of ordinary audiovisual recording. The candidate’s metamorphosis into a tiger and the physical destruction of an adversary converted into a visual object condense leadership, strength, and rupture within a single image. AI thus operates as a technology of narrative intensification: it materializes political metaphors, dramatizes antagonisms, and allows the candidate to visually occupy the position of protector or victor. The campaign translates complex political problems into audiovisual representations of threat, rescue, order, and national belonging. Certainty here emerges less from extensive programmatic explanation than from a visual organization of conflict that distributes recognizable positions between threat, adversary, and the capacity to respond.

Research on synthetic political videos has shown that these technologies can alter the conditions through which audiences evaluate the authenticity of content (Vaccari & Chadwick, 2020). The tiger piece, however, represents a different mode of intervention: its fantastic character is visible, and its political force depends less on its capacity to be mistaken for documentary footage than on its ability to transform an electoral narrative into a recognizable spectacle. Recent studies suggest that the communicative effects of synthetic content also depend on the emotions it activates and the interpretive frames it makes available to audiences (Hameleers, 2024). This distinction is central to the case examined here because generative AI expands the propagandistic repertoire through images that can be manifestly fictional while retaining a strong capacity for affective orientation.

The use of generative artificial intelligence in these pieces can be understood as a cultural mediation that turns political antagonisms into spectacular and emotionally accessible visual scenes. The synthetic image reorganizes the sensible form of the message by condensing threat, protection, and victory into representations designed for rapid recognition and circulation. In the case examined here, this operation contributes to producing an affective reality in which the country appears under threat, the candidate occupies the position of protector, and the adversary becomes associated with moral risk or political disorder.

This visual messaging also extended beyond the candidate’s official accounts. Vincent Ramos, a digital activist close to De la Espriella and part of the candidate’s network, participated in circulating a narrative that linked Gustavo Petro’s government, Armando Benedetti, and Paloma Valencia’s campaign to an alleged operation intended to block De la Espriella (Ortiz, 2026). In a piece circulated in early May, Ramos articulated this interpretation through AI-generated imagery and claims that represented political adversaries as participants in covert coordination; a few days later, on 7 May, he further intensified this narrative with the post “URGENT: Petro and Benedetti’s entire plan with Paloma’s campaign,” whose lack of clear evidentiary support was noted in journalistic coverage (Stacey, 2026; Sánchez Nieto, 2026). These pieces are treated here as supportive circulation produced by an actor who is connected to the candidate rather than as official campaign content. Their relevance lies in showing how repertoires of threat and enemy construction could extend beyond the candidate’s accounts, be reformulated through synthetic visualities, and recirculate through actors capable of intensifying suspicion, indignation, and antagonism.

The campaign’s affective repertoire also included modulations of euphoria, antagonism, and belonging that expanded the relationship between threat and protection. At the campaign-closing event in Barranquilla on 23 May, a tiger formed by drones, fireworks, music, and Caribbean symbols transformed political identification into a collective experience of spectacle, celebration, and strength (Osorio, 2026b). In another case, the message “THEY ARE LEAVING! The regime and the usual ones,” circulated on 29 May, organized antagonism through a language of expulsion and anticipated victory that transformed electoral competition into the defeat of those represented as responsible for the country’s deterioration (¡YA SE VAN!, 2026). Belonging also acquired territorial expressions: the appeal “Costeño vota costeño” (“Caribbean voter, vote Caribbean”) connected regional identity, homeland, and mobilization through an invitation to join the “Army of Defenders of the Homeland” (“Costeño vota costeño”: Abelardo de la Espriella llama a los samarios a respaldar su candidatura, 2026; De la Espriella, 2026). These expressions do not constitute repertoires independent from those already examined; they show how spectacle, indignation, and belonging were able to intensify the same affective sequence linking threat, authority, protection, and community.

Patriotic visuality reinforced this production of certainty. The nation was presented as a vulnerable community, a disputed territory, and a moral inheritance called upon to be defended. This representation resonates with a repertoire widely documented in contemporary nationalisms. Digital environments participate in the reproduction of national belonging and can foster more fragmented and polarized imaginaries of political community (Mihelj & Jiménez-Martínez, 2021). Within far-right populist nationalisms, homeland can also acquire an intense affective dimension when represented as a familiar and rightful space whose continuity is under threat (Shoshan, 2026). The campaign turned national symbols into resources of everyday identification through flags, colors, slogans, gestures of authority, and appeals to the defense of the homeland. The particularity of De la Espriella’s case lies in the way generative AI expands this repertoire by turning the defense of the nation into a visualizable scene. The tiger destroying “Los de Siempre” materializes a fantasy of recovery in which threat, territory, and protective leadership can be condensed within a single synthetic representation. In this register, the homeland functions as a moral image that organizes the relationship between citizenship, order, and belonging. This patriotic visuality provides an empirical expression of the algorithmic homeland: established repertoires of national protection acquire contemporary forms of staging through digital formats, platform circulation, and generative visuality.

The religious dimension introduced an additional form of legitimation. Religion operates throughout the campaign as a transversal affective repertoire that organizes uncertainty through languages of faith, protection, mission, and moral certainty. Appeals to God, faith, and spiritual protection make it possible to instrumentalize uncertainty as an experience of threat and salvation. On the Westcol stream, this messaging appeared through the display of the Bible, the conversion narrative, and the candidate’s identification with David facing Goliath; elsewhere in the campaign, references to God, prayer, and a “spiritual battle” extended this providential reading of the electoral contest. Leadership appears invested with a task of defense, the community recognizes itself as the bearer of higher values, and the adversary becomes incorporated into a narrative of threat. Technological mediation shapes the form taken by this repertoire: streaming brings it closer to biographical confession and a scene of authenticity, while audiovisual posts condense it into symbols, slogans, and rapidly circulating images.

Security constitutes another affective core of the campaign. Messages of order, authority, and control condense dispersed experiences of fear and frustration, as illustrated on the Westcol stream by the combination of images of armed groups, the characterization of insecurity as a “pandemic,” and the promise to exercise authority with an “iron hand.” Insecurity thus takes the form of a demand for guidance, while the candidate is imbued with expectations of protection and capacity for response. This repertoire crosses different campaign formats and acquires particular intensity when generative technologies turn it into visual scenes of confrontation, destruction of the adversary, and restoration of order. AI contributes narrative plasticity here by making it possible to represent, through impossible images, a political relationship already present in the discourse between threat, protective force, and the recovery of the homeland. The emotional clarity of these compositions favors their adaptation to brief, visual, and readily recirculable platform content.

The coexistence of these repertoires allows the campaign to address different political sensibilities. The differentiation observable in the corpus lies in the affective languages and public forms of address that acquire force according to format and moment of circulation. Religious appeals provide frames of certainty and moral protection; security messages articulate fear, authority, and order; references to breaking with traditional politics mobilize antagonism and renewal; and the candidate’s entrepreneurial persona connects political leadership with efficiency, success, and managerial capacity. These variations retain a common promise of restoration while expressing it through different affective languages. The campaign thus acquires an adaptable texture in which homeland, security, religion, rupture, and authority can be combined according to format, event, and context of circulation.

Generative AI strengthens this logic through the possibility of producing visual variations in a recognizable political repertoire. A slogan can become an epic video, a religious appeal, an image of protection, a promise of security, a scene of national defense, and a critique of the adversary as a threatening representation. The plasticity of these technologies makes it possible to recombine symbols, characters, and antagonisms and adapt their expression to different formats, campaign moments, and contexts of circulation. In democratic terms, Jungherr (2023) situates artificial intelligence as a mediation that transforms information, public communication, participation, and power relations. In the case examined here, this transformation is expressed in the capacity to multiply and reformulate political representations through a high degree of symbolic adaptability, making visually available metaphors, threats, and promises that previously circulated within the campaign’s discourse.

The articulation of these mediations and affective repertoires reveals a recurrent relationship across the corpus. Streaming organizes forms of proximity and performative authenticity; generative AI expands the visual plasticity of antagonism; and patriotic, religious, and security repertoires connect belonging with threat, moral certainty, and protection. From this configuration emerges the algorithmic homeland as a national imaginary in which the country is represented as a threatened and morally ordered community whose restoration requires an authority capable of protecting it. Its algorithmic dimension refers to the conditions of visibility, recommendation, circulation, and recirculation through which these representations acquire intensity within platform environments. The concept therefore helps explain how established repertoires of homeland, threat, and protection are reformulated through the sociotechnical conditions of contemporary electoral campaigning.

4. Platform Democracy and the Political Implications of Algorithmic–Affective Campaigning

The case examined here makes it possible to move from the specific campaign materials toward the democratic implications of a communicative environment in which platform visibility, affective intensity, and technological mediation are closely related. De la Espriella’s digital strategy shows how political authority can be staged through formats that combine proximity, spectacularization, symbolic plasticity, and recognizable promises of protection and restoration. The candidate’s entrepreneurial persona also introduces a repertoire of efficacy, decisiveness, and managerial competence, integrated here as a secondary dimension within a broader platformized campaign configuration. From this perspective, the discussion focuses on how platform infrastructures and generative media participate in reorganizing the conditions of political visibility, identification, and public judgment.

Digital platforms participate in this configuration by organizing the circulation of content through interaction metrics, formats, and visibility rules. Platformization transforms cultural production through the incorporation of the economic and infrastructural extensions of platforms into the circulation of symbolic goods (Nieborg & Poell, 2018). In electoral communication, this logic places campaign materials within environments where their exposure depends on formats, temporalities, recommendation systems, and practices of recirculation. In the corpus examined here, these conditions can be observed in the coexistence of live broadcasts, short-form audiovisual pieces, and fragments that continue to circulate after the initial event. Electoral content acquires value when it circulates, produces impact, and becomes appropriable by communities that comment on, reinterpret, and redistribute it through their own networks.

Algorithmic–affective campaigning also displays formal affinities with some logics of commercial communication, particularly in the construction of a recognizable political figure, the repetition of symbols, the adaptation of content across formats, and the production of readily identifiable promises. In De la Espriella’s case, the figure of “El Tigre,” the image of the successful entrepreneur, and repertoires of authority, security, and restoration contribute to turning political attributes into condensed signs of leadership. This dimension helps explain how candidate personalization participates in constructing the candidate’s campaign promises within environments structured by attention and visibility. Its interpretive scope, however, remains at the level of the publicly observable forms of representation and circulation contained in the corpus. The evidence examined makes it possible to analyze how the campaign presents, combines, and recirculates these promises; it does not establish the internal procedures of data exploitation, behavioral profiling, or informational extraction through which they may have been distributed.

The candidate’s entrepreneurial persona introduces an additional repertoire of legitimacy based on efficacy, decisiveness, and managerial competence. De la Espriella projects attributes associated with his private-sector trajectory onto the representation of political leadership, presenting speed, administration, growth, and control as signs of governing capacity. This operation also appears on the Westcol stream, where the electoral contrast incorporates references to business performance and economic growth as criteria for differentiating political projects. The outsider position thus acquires a managerial dimension: distance from professional politics is combined with an image of authority supported by entrepreneurial experience and capacity for execution. Within algorithmic–affective campaigning, this repertoire becomes intertwined with security, restoration, and rupture, reinforcing the representation of a leadership capable of producing results.

This translation of entrepreneurial attributes into the political sphere has implications for the way government is represented. Institutional complexity can be communicatively condensed into personal qualities such as decisiveness, efficacy, and capacity for command, shifting the center of the promise toward the candidate. Authority therefore acquires a performative dimension in which trajectory, style, and capacity for execution operate as resources of legitimacy. This personalization converges with the dynamics observed in streaming and generative visuality: different formats present De la Espriella as representing promises of protection, transformation, and restoration that structure the campaign’s political imaginary.

Research on digital persuasion provides a broader context for interpreting these transformations. Previous studies have shown that messages tailored to psychological traits inferred from digital traces can increase persuasive effectiveness in particular contexts (Matz et al., 2017). In the corpus examined here, however, the evidence supports a more limited claim: the campaign circulated different affective repertoires through varied formats and communicative moments, while the materials analyzed do not establish whether these variations resulted from psychological profiling, behavioral segmentation, or personalized distribution systems. The ethical relevance of the case therefore lies in the public conditions of mediation, visibility, and circulation that structure electoral experience.

The discussion of digital manipulation remains relevant as a normative framework for distinguishing political persuasion from forms of influence that exploit vulnerabilities through opaque mechanisms (Susser et al., 2019). In this study, that distinction functions as an analytical boundary rather than as a description of a demonstrated campaign practice. The observable concerns relate primarily to content provenance, the degree of artificial intervention, and the conditions through which synthetic representations circulate as part of electoral discourse. Transparency therefore acquires democratic relevance because it makes it possible to assess who produced a piece, which elements were generated or modified, and under what conditions it enters public conversation.

Generative artificial intelligence expands this concern by introducing a high degree of plasticity into the production of political scenes. A campaign can multiply versions of a narrative, recombine symbols, and modify their expression across formats, campaign moments, and contexts of circulation. The ethical question extends toward the regime of public sensibility formed through synthetic content, rapid dramatizations, and narratives designed to produce reaction. Synthetic political videos can increase uncertainty about the authenticity of content and affect trust in information, especially when they circulate in environments of high polarization and low public verification (Vaccari & Chadwick, 2020). In the corpus examined here, this problem also takes another form when pieces openly display their fantastic character. The image of a candidate transformed into a gigantic tiger derives its political force from the dramatization of antagonism and the succinct visual representation of a promise of victory, whereas other forms of synthetic content may introduce greater ambiguity regarding provenance or artificial intervention. In democratic terms, both situations make relevant the conditions of transparency and contextualization through which publics can recognize how a political representation has been produced and evaluate the interpretive frame it advances.

Platform democracy can therefore be understood as an ambivalent communicative environment. It expands opportunities for participation, allows new actors to enter political conversation, and partially displaces traditional control over public visibility. At the same time, this participation takes place within architectures that organize exposure through formats, interaction metrics, recommendation systems, and commercial logics. On the Westcol stream, for example, chat, reactions, excerpts, and recirculation become part of the political event, while the platform infrastructure establishes the conditions under which these interventions acquire visibility and continuity. The citizen thus participates as interlocutor, audience member, and agent of circulation within an environment whose technical conditions contribute to organizing public experience.

This ambivalence makes it possible to understand the digital campaign as an electoral practice situated within the specific cultural and technical conditions of platforms. Its communicative capacity lies in connecting social uncertainties with recognizable promises of protection, belonging, authority, and restoration through formats capable of circulating across different digital environments. Within this framework, the algorithmic homeland makes it possible to trace how the affective repertoires identified in the analysis enter circuits of political attention shaped by formats, recommendation systems, temporalities, and practices of recirculation.

Technology companies form part of the infrastructural environment in which this communication takes place. Kreiss and McGregor (2018) have shown how companies such as Facebook, Twitter, Google, and Microsoft actively participated in shaping electoral communication during the 2016 U.S. presidential cycle. This precedent helps situate the Colombian case within a broader ecosystem of technological intermediation, although it does not constitute evidence that these companies or their services performed equivalent functions within the campaign examined here. Political teams, content creators, and users produce, interpret, and recirculate messages, while platform infrastructures establish rules, metrics, and recommendation systems that condition their visibility in differentiated ways.

The democratic implications of this configuration concern the conditions under which political judgment is formed and circulated. The immediacy of streaming favors forms of presence and public evaluation constructed in real time; fragmentation extends the event through clips and appropriations; and synthetic visualities introduce new questions of provenance, contextualization, and recognition of artificial intervention. These dynamics do not, in themselves, demonstrate a loss of deliberation or a general erosion of public truth, but they show that the evaluation of candidates and political conflicts increasingly takes place within environments where proximity, spectacularization, speed, and technological mediation participate in the production of visibility.

De la Espriella’s campaign therefore reveals a reconfiguration of the relationship between political communication, technology, and affect. In this discussion, the algorithmic homeland synthesizes the empirical pattern identified throughout the analysis: established repertoires of homeland, threat, and authority acquire new forms of visibility, intensity, and circulation through streaming, generative visualities, and platform infrastructures. Its analytical relevance lies in specifying the mediating conditions through which these repertoires are staged, appropriated, and redistributed, while locating political agency in the actors, practices, and infrastructures involved in this process.

5. Conclusions

Abelardo de la Espriella’s digital campaign reveals a mode of electoral communication in which technological mediations and affective repertoires operate in concert. The concept of algorithmic–affective campaigning describes this mechanism as a vocabulary of production, visibility, and circulation in which platform formats, digital temporalities, and recommendation systems interact with languages of threat, protection, certainty, belonging, and authenticity. The contribution of the case lies in showing how these dimensions can acquire coherence within a strategy combining streaming, audiovisual content, and generative visuality without reducing the explanation to technology itself.

Streaming and generative artificial intelligence performed differentiated functions within this configuration. The Westcol stream transformed the candidate’s prolonged exposure into a scene of proximity and performative authenticity in which religion, security, and authority acquired biographical and interactive dimensions. Generative AI, in turn, expanded the campaign’s visual possibilities by materializing political metaphors and transforming antagonisms into synthetic scenes of confrontation, strength, and restoration. The figure of the tiger provides a particularly clear example: its communicative force lies in visually condensing a narrative already present in the political discourse, in which threat requires a force capable of defeating it and restoring order.

From this articulation emerges the algorithmic homeland as a national imaginary in which Colombia is represented as a threatened and morally ordered community in need of protection. The concept does not suggest that algorithms autonomously produce the nation or that repertoires of homeland, enemy, and restoration are historically new. Its analytical value lies in identifying how these repertoires acquire contemporary forms of staging, intensification, and circulation within platform infrastructures and through technologies capable of generating new political visualities. In this respect, the Colombian case resonates with transformations observed in other Latin American experiences, including the intensive use of affective repertoires and digital environments in Javier Milei’s campaign in Argentina and dynamics of polarization and participation documented in recent campaigns in Brazil, Chile, and Peru (Montero, 2024; Fenoll et al., 2024). De la Espriella’s case is distinctive in the particularly visible combination of streaming, influencer mediation, patriotic and religious repertoires, and generative artificial intelligence.

These forms of communication raise democratic questions concerning the conditions through which political content acquires visibility and can be publicly evaluated. Synthetic visualities make provenance, contextualization, and recognition of artificial intervention particularly relevant, while platforms organize exposure through formats, metrics, recommendation systems, and practices of recirculation. These conditions reshape the communicative environment within which citizens and publics interpret candidates, threats, and political promises, although this study does not establish causal effects on voting behavior or determine how content was technically distributed to specific groups.

The scope of these conclusions should be understood within the limits of the qualitative design adopted. The core analytical corpus consists of one complete stream and six digital pieces selected for their relevance to the mediations and repertoires under study; the findings therefore support an interpretation of discursive, visual, and affective mechanisms rather than an exhaustive reconstruction of the campaign, measurement of persuasion, demographic profiling of audiences, or demonstration of internal microtargeting strategies. Future research could compare these vocabularies across other Latin American campaigns, systematically examine the reception of synthetic political content, and explore how different platform infrastructures reorganize nationalist, religious, and security repertoires. From this perspective, algorithmic–affective campaigning and the algorithmic homeland offer concepts for examining transformations in electoral communication in which technology, affect, and national imaginaries converge within platform democracy.