Preliminary Report ahead of 2026 General Elections in Spain

---

---

Context, scope & approach

Election Day is set for Sunday, November 29, 2026. The Rapid Response System (RRS) activates during election cycles four weeks before the vote and remains active one week after, running from November 2 to December 6, 2026.

The analysis covers developments in Spain’s information landscape between 2023 and 2026. The audit includes: 

1. A quantitative review of 8,086 fact-checks from 2023–2026 tagged by five fact-checking organizations in the IBERIFIER Plus repository (EFE Verifica, Infoveritas, Maldita.es, Newtral, and Verificat), out of 9,676 total records logged between 2017 and 2026; 
2. Tracking of content circulation across open, high-reach networks (X, representing 36% of tracked platform mentions, and Facebook, at 24.3%), short-form video platforms (TikTok, 10.5%, and Instagram, 5.5%), and private messaging apps (WhatsApp, 6.5%, and Telegram, 2.1%); 
3. Evaluation of coordinated FIMI operations (Doppelgänger and Portal Kombat, with documented cases in Spain, alongside Russian networks operating across Europe such as Storm-1516), generative AI and LLM grooming, impersonation of news outlets (El País, El Mundo, 20 Minutos, RTVE) and public institutions, and activity from pseudo-media sites and high-reach accounts.

Platform percentages are calculated based on total origin mentions, rather than the total number of fact-checks, as a single entry may cite multiple platforms.

Figure 1: X and Facebook account for six in ten platform mentions

Platform	Mentions	%
X (Twitter)	5,267	36
Facebook	3,558	24.3
Tiktok	1,534	10.5
WhatsApp	9444	6.5
Instagram	800	5.5
Telegram	312	2.1
Threads	152	1
Youtube	116	0.8
Others	1,930	13.2
IBERIFIER repository, in IBERIFIER, 2026a, fig. 29 · 14,613 source mentions Source: IBERIFIER Repository, compiled in Badillo et al. (2026, Fig. 29).

The analytical methodology combines four levels of cross-referencing: 

1. Incident categorization using the TTP (Tactics, Techniques, and Procedures) taxonomy integrated into the hub’s fact-checking infrastructure; 
2. Qualitative case studies on electoral delegitimization (2023 general election and 2024 European election) and disinformation during the October 2024 Valencia DANA floods; 
3. Social perception indicators from CIS Study 3421 (September–October 2023, n = 27,433), the Eurobarometer, and the Digital News Report; and 
4. Institutional reports from the Department of National Security (DSN) and the European External Action Service (EEAS).

This report has several limitations. The repository represents a targeted sample: fact-checking organizations select items based on virality, public demand, and editorial judgement, meaning it does not capture all circulating content. WhatsApp and Telegram are underrepresented due to technical access barriers. Survey data reflects self-reported user perceptions. Additionally, the data originates from the IBERIFIER Plus repository compiled by that project’s consortium; maintaining full transparency regarding data sources underpins the independence of this analysis (IBERIFIER, 2026a, pp. 12–15).

Diagnosis of the information ecosystem in Spain

The pre-electoral information environment reflects shifts in news credibility and media gatekeeping. Overall trust in news stands at 33%, two points above 2025 (Digital News Report España 2026). In the digital sphere, 74% of the population say they are worried about telling what is true from what is false online, five points above 2025 and the highest level in the series (Digital News Report España 2026). Among youth aged 15 to 24, 80% report encountering disinformation regularly, while 13% systematically verify the news they consume (Evercom, 2025). Distrust toward press and radio shows no linear ideological correlation, though it reaches higher levels at the ends of the political spectrum compared to the center (CIS, Study 3421).

Figure 2: Three in four worry about true vs false online; a third trust the news

Indicator	Data	Source and Year
Trust in news	33%	Digital News Report España 2026; 2 points above 2025
Concern about telling true from false online	74%	Digital News Report España 2026; highest in the series
News fatigue	44%	Digital News Report 2024 (Sierra, 2024a)
Active news avoidance	37%	Digital News Report España 2026; 1 point above 2025
“News NEETs” (low trust and low interest)	35%	Digital News Report España 2026; 53% among under-25s
Source: Digital News Report España 2026 (UNAV, 2026); fatigue from DNR 2024 via IBERIFIER, 2026a. Fatigue from IBERIFIER, 2026a, pp. 77–87; trust, concern, avoidance and news NEETs from Digital News Report España 2026.

News consumption habits in Spain reflect these trust dynamics. Overall, 44% of citizens report news fatigue—a figure rising to 51% among women and 38% among men. Active news avoidance stands at 37%, one point above 2025, and reaches 44% among 25-to-34-year-olds (Digital News Report España 2026). People with low trust and low interest in news (“news NEETs”) are 35% of the population and 53% of under-25s (Digital News Report España 2026). Only 19% of adults under 24 trust traditional media, shifting their news access from mainstream outlets to short-form video platforms like TikTok, Instagram, and YouTube. Algorithmic distribution now accounts for 66% of digital news consumption, driven by content creators, podcasters, and high-reach accounts overtaking traditional editorial gatekeeping.

A global Spanish-speaking population exceeding 640 million people makes Spain a key target for influence operations

Compounding these shifts is a drop in web referral traffic. The deployment of AI-generated search summaries by default (Google AI Overviews, operational since January 2026) has curbed traffic to digital newspapers. In February 2026, GfK DAM data showed El Mundo lost two million unique visitors and experienced a 23% month-over-month drop in pageviews; El Español saw a 27% year-over-year decline in pageviews, while La Vanguardia saw its traffic halved. This drop impacts investigative journalism revenues precisely as the cost of generating synthetic disinformation falls. Technologically, generative AI is deployed in audiovisual deepfakes, voice cloning of public figures, and algorithmic manipulation (LLM grooming), detailed further in Section 3 (IBERIFIER, 2026, pp. 80–91).

Spain’s vulnerability profile highlights an overlap between political instability, cybercrime, and foreign influence operations (FIMI). Cyber-enabled threats include domain and brand impersonation of major media outlets (El País, El Mundo, 20 Minutos, RTVE) and public institutions, used to spread political narratives and cryptocurrency scams featuring manipulated images of public figures.

Meanwhile, a global Spanish-speaking population exceeding 640 million people (Instituto Cervantes, 2026) makes Spain a key target for influence operations: according to the European External Action Service (EEAS), Kremlin-linked disinformation networks operate across Spain and Latin America, leveraging RT and Sputnik via diplomatic accounts and proxy networks; content originating from campaigns targeted at U.S. Hispanic voters frequently reappears in Spain with minimal modifications (IBERIFIER, 2026a, pp. 102–104).

Technological vectors, tactics, and specific vulnerabilities

The shared Spanish language acts as a primary vector for online disinformation. With over 640 million speakers worldwide (Instituto Cervantes, 2026), the lack of language barriers enables rapid cross-border circulation of content between the United States, Spain, and Latin America (Salaverría, 2026). Fact-checking organizations have documented narratives on migration, crime, and voter integrity initially designed for U.S. Hispanic audiences appearing in Spain shortly thereafter. The sources consulted do not document Russian networks deliberately testing narratives in one country in order to move them to another. What is documented is that content moderation resources for non-English languages remain limited: across X and Pinterest, only 8% of moderators speak an official non-English EU language, while internal Facebook documents from 2021 revealed that 87% of its anti-disinformation budget was allocated to English content, spoken by only 9% of its user base (IBERIFIER, 2026a, pp. 102–104).

Narratives targeting Spanish audiences focus on undermining trust in democratic processes, European Union institutions, and immigration management. Ahead of the 2024 European elections, EU-related disinformation accounted for 15% of all false content detected in the month preceding the vote, according to EDMO data cited by the European Board for Digital Services (IBERIFIER, 2026, p. 128).

Narratives targeting Spanish audiences focus on undermining trust in democratic processes, Eu institutions, and immigration management

Documented Foreign Information Manipulation and Interference (FIMI) incidents in Spain stem primarily from two Russian-linked networks. Doppelgänger cloned European news websites using domain networks and redirects to bypass platform bans, alongside purchasing Meta ads to amplify fake articles. Meanwhile, Portal Kombat, operating under the Pravda brand, launched regional portals tailored to Galician, Basque, and Catalan audiences (IBERIFIER, 2026a, p. 107).

During the 2023 General Election, the EEAS identified pro-Kremlin Telegram channels acting as core distribution nodes; attackers also cloned the official Madrid regional government website to post a fake bomb threat attributed to ETA at polling stations (pp. 123–125). Security reports logged over 40 Russian operations targeting the 2024 European Parliament elections (p. 127).

Generative AI has expanded the scale and sophistication of disinformation (IBERIFIER, 2026b). Audio cloning and video manipulation have transitioned from geopolitical targets (such as the deepfake video call to the Mayor of Madrid) to financial fraud, with manipulated videos of the Prime Minister and RTVE journalists used to promote investment scams. In the IBERIFIER repository, fact-checks involving AI rose from ~2% monthly in 2023–2024 to 18.5% in early 2026 (Fig. 25) (IBERIFIER, 2026a, pp. 105–106).

Synthetic satire, partisan communication, and the blurring of electoral risk

Ahead of the electoral period, there is growing use of generative artificial intelligence tools by political parties, content creators, and affiliated accounts to produce synthetic audio, images, and video aimed at satirical critique, caricature, and disparaging political opponents. Although the primary intent of these contents usually falls within the domain of political parody and is recognized as such by the audience, their unlabeled proliferation introduces significant vulnerabilities into the information ecosystem.

The widespread dissemination of deepfakes and synthetic media for partisan confrontation blurs the line between freedom of expression and subtle manipulation, accustoms the electorate to consuming unverified hyperrealism, and erodes trust in the authenticity of genuine documentary evidence. Furthermore, this synthetic saturation fuels the so-called “liar’s dividend,” enabling political actors to proactively contest the authenticity of real recordings or evidence by claiming they were fabricated using artificial intelligence.

Alongside media manipulation, generative AI poses a threat through LLM grooming—the automated publishing of multilingual fake articles designed to corrupt training data for large language models (Larraz & Salaverría, 2026). A recent IBERIFIER report (2006a) indicates networks like Portal Kombat and CopyCop have initiated these tactics (p. 106). When users query models about candidates, voting procedures, or party platforms, systems may regurgitate these implanted biases. To date, published research has not confirmed specific LLM grooming cases in Spanish, making it an identified threat vector rather than an established occurrence.

Media typosquatting and the convergence of political disinformation with cybercrime represent further vulnerabilities. By registering lookalike domains and employing redirects, actors distribute deceptive narratives disguised as legitimate journalism. Between 2023 and 2026 (n = 8,086), the most frequently impersonated entities in fact-checks were the National Police (149), El País (61), the Tax Agency (54), Cuatro (50), and the Civil Guard (45); El Mundo registered 26 instances, and RNE/RTVE/TVE combined for 20 (Fig. 26). Fraud networks deploy altered images of public officials, ministers, and journalists to funnel users into cryptocurrency schemes, leveraging paid verification badges on social platforms to establish credibility. Although the proportion of pure financial fraud in the repository fell from a peak of 20.8% in June 2022 to single digits by 2024, crypto scam fact-checks remain consistent at 15 to 24 cases annually (6 in 2022, 24 in 2023, 15 in 2024, and 18 in 2025; Figs. 27 and 28) (IBERIFIER, 2026, pp. 107–111).

The distribution and amplification of these tactics rely on pseudo-media outlets, high-reach accounts, and cross-platform sharing. Pseudo-media sites mimic legitimate news organizations while ignoring journalistic standards, serving as hubs for hyper-partisan content driven by ideological goals or ad-tech monetization (Palau-Sampio, 2023). Their reach expands through high-profile accounts that amplify their stories.

During the DANA flood crisis, public figures and digital influencers shared unverified claims under the premise of revealing truths suppressed by official channels. This tactic matches what the literature calls diagonalism: a strategy that breaks down the ideological boundaries between left and right to capture social discontent from different angles, and which far-right actors use to redirect historically progressive discourses (critique of globalization, skepticism toward elites) toward nationalist, xenophobic, and conspiratorial narratives (IBERIFIER, 2026a, pp. 107–111).

Distribution patterns vary significantly by platform type. Open social networks account for the majority of tracked mentions: X leads at 36%, followed by Facebook at 24.3%. X shifted from 36% in 2023 to 38% in 2024, 33.5% in 2025, and 37.3% in preliminary 2026 data (Fig. 30). TikTok (10.5%) and Instagram (5.5%) make up smaller shares of the total repository, though 46.6% of youth aged 15 to 24 identify TikTok as their primary vector for misleading content (Evercom, 2025). WhatsApp (6.5%) and Telegram (2.1%) function as closed or semi-closed environments without public content moderation. Telegram’s low share reflects technical monitoring constraints; forensically, it acts as an initial distribution node for networks tied to foreign interference. Telegram is not designated a Very Large Online Platform (VLOP) under the DSA and falls outside the HODIO monitoring program, which covers Facebook, Instagram, TikTok, X, and YouTube (IBERIFIER, 2026a, pp. 112–113). Set up in March 2026, HODIO is the Spanish Observatory on Racism and Xenophobia, which operates under the Ministry of Inclusion, Security Social and Migration.

Regulatory framework and public policies

Spain’s response to disinformation and foreign interference combines European Union regulations with national frameworks. At the EU level, the Digital Services Act (DSA) mandates that Very Large Online Platforms (VLOPs) and Search Engines (VLOSEs) assess and mitigate systemic risks to electoral integrity, freedom of expression, and public discourse. This is reinforced by the European Media Freedom Act (EMFA), designed to safeguard editorial independence, media ownership transparency, and state advertising allocation. Article 18 of the EMFA establishes a right of prior notification: very large platforms must notify independent media outlets before restricting their content (IBERIFIER, 2026a, p. 97).

The EU Artificial Intelligence Act (AI Act) sets out transparency requirements for synthetic content under Article 50: generative AI providers must watermark output in machine-readable formats, and deployers must disclose deepfakes involving real people, places, or events. These rules took effect on August 2, 2026. Providers of systems deployed prior to that date have until December 2, 2026, to comply with watermarking standards (Fillon et al., 2026). The November 1–December 6 election monitoring period falls within this transition window. To date, synthetic media detection by fact-checkers and law enforcement has relied on third-party forensic tools and voluntary platform labeling (IBERIFIER, 2026a, p. 105).

Regulation (EU) 2024/900 on the transparency and targeting of political advertising (TTPA) governs digital political ads. On July 25, 2025, Meta announced it would pause political advertising sales in the EU starting October 2025 due to compliance challenges under the new rules; Google announced a similar policy shift (Iyer, 2025). This regulatory shift risks driving political spending toward unmonitored channels, such as sponsored influencer campaigns, though available data has not yet verified this pivot.

Nationally, Spain coordinates anti-disinformation measures through its Action Procedure Against Disinformation (Order PCM/1030/2020). Managed by the Permanent Commission against Disinformation and the Department of National Security (DSN), the framework establishes four operational levels (prevention, detection, alert, and response) connected to the EU Rapid Alert System (RAS) and the European Cooperation Network on Elections. The RAS is distinct from the RRS described in Section 6: the former connects national contact points and EU institutions, whereas the latter bridges independent researchers, fact-checkers, and online platforms. During elections, the Electoral Security Network deploys a working group chaired by the DSN; the Permanent Commission activated its protocol during the 2023 and 2024 elections (Badillo et al., 2026, pp. 93–94, 122). This structure is complemented by the Forum Against Disinformation Campaigns (Foro contra las campañas de desinformación en el ámbito de la Seguridad Nacional), established in May 2022, comprising six government officials and ten representatives from civil society, media, academia, and industry (p. 96).

Policy initiatives expanded with the Action Plan for Democracy approved in September 2024 (comprising eleven focus areas and thirty-one measures) alongside the drafting of a National Strategy Against Disinformation Campaigns. Announced by the National Security Council in January 2025, the strategy remained unpublished as of September 2026. These policies target media ownership transparency, state advertising regulation, and manipulated content online.

In 2026, the government launched the HODIO program to monitor hate speech. In the DSN’s latest risk perception report, disinformation ranks second, behind cyberspace vulnerabilities (IBERIFIER, 2026a, pp. 94–101, 120).

Independent observatories like IBERIFIER provide empirical data to evaluate the impact of these legal frameworks across digital channels.

Electoral antecedents and recent crises

Analysis of the July 23, 2023 general election and the June 2024 European Parliament elections shows recurring patterns of institutional delegitimization where foreign interference infrastructures (FIMI) intersected with domestic political polarization. Regional elections in 2023 and 2024 (Galicia, Basque Country, and Catalonia) saw lower levels of interference, though local patterns were observed and the Permanent Commission activated its monitoring protocols for each contest (IBERIFIER, 2026a, p. 122).

In 2026, events such as the migration crisis in Ceuta and the fatal high-speed train accident in Urda have triggered disinformation waves

July 2023 General Elections

According to Maldita.es and Democracy Reporting International, 24% of debunked content in the month preceding the vote targeted electoral integrity, rising to 74% on Election Day. As detected in other countries’ elections (Salaverría, Rodríguez Virgili & Sierra, 2026), primary narratives questioned mail-in voting—generating 21% negative social media volume—and spread unsubstantiated claims alleging election technology vendor Indra manipulated vote tallies.

In the IBERIFIER repository, electoral fraud claims formed the single largest category during the campaign (19 fact-checks). In the final 72 hours before the vote, attackers spoofed the official Madrid regional government website to post a fake ETA threat at polling places, while claims of railway sabotage on the Madrid-Valencia line circulated on Election Day to suggest voter suppression. Following the vote, inauthentic networks and Russian proxies, among other actors, spread claims that the EU, NATO, Morocco, and the Spanish Crown had forced the outcome (IBERIFIER, 2026a, pp. 123–126).

June 2024 European Parliament Election

The repository logged 645 fact-checks between May and July 2024. Anti-immigration claims formed the largest cluster (32), followed by health and vaccines (14), NATO and security (12), and the economy (9); election fraud claims accounted for 6 entries. A prominent incident involved a recycled 2019 video, presented as a call to vote for the governing party under the threat that social benefits for Moroccan migrants would end; X subsequently flagged the video as manipulated media. Security services logged over 40 Russian-backed operations targeting the EU elections, though the European Commission concluded these efforts failed to alter voting outcomes. Maldita.es evaluated over 1,300 posts previously debunked by European fact-checkers and found platforms took no visible action on half of them; YouTube failed to act on 75% of reported cases and X on 70% (IBERIFIER, 2026a, pp. 127–129).

October 2024 Valencia DANA Flood Crisis

Flash floods on October 29, 2024, claimed 230 lives and impacted over 300,000 residents. The repository cataloged 83 crisis-related fact-checks between October and December: 56.6% rated “false” and 33.7% “explanatory.” By topic, 56.6% involved environmental disasters, 18.1% emergency alerts, and 15.7% scientific claims. Disinformation focused heavily on state agencies: the State Meteorological Agency (AEMET) and the Júcar Hydrographic Confederation were accused of concealing forecast data or delaying alerts; false claims alleged the government was demolishing dams; and numerous content creators claimed without evidence that hundreds of bodies lay in an underground car park of a shopping centre, later disproved by National Police inspections (Sánchez Torres, Pastor, & Magallón Rosa, 2025). Other narratives pushed weather modification theories, used clips from a television series to suggest that elites had planned the tragedy, falsely accused immigrants of looting, and attacked relief efforts by the Red Cross and Cáritas. According to Maldita.es, false TikTok videos garnered over 120 million views without platform intervention. An IBERIFIER report (2026a, pp. 130–133) warns that the real reach of these narratives is under-represented, because fact-checkers work on open networks and not on private Telegram groups or WhatsApp broadcasts.

Figure 3: Each episode has its own dominant narrative

Episode	Narrative or Category	Fact-Checks
2023 General Elections (June–August)	“Pucherazo” (Electoral fraud)	19
2023 General Elections	Indra vote counting fraud	6
2023 General Elections	Mail-in vote tampering	6
2023 General Elections	Anti-immigration	5
2024 European Elections (May–July)	Anti-immigration	32
2024 European Elections	Health and vaccines	14
2024 European Elections	NATO and security	12
2024 European Elections	Economic anxiety	9
2024 European Elections	Electoral fraud or manipulation	6
2024 DANA Crisis (October–December)	Environmental disasters	47
2024 DANA Crisis	Emergency alerts	15
2024 DANA Crisis	Science	13
2024 DANA Crisis	Politics	6
IBERIFIER repository, in (IBERIFIER, 2026a, figs. 39, 42 and 45). Source: IBERIFIER Repository, compiled in IBERIFIER, 2026a, Figs. 39, 42, and 45. General and European election tallies were compiled using title keyword searches and contain overlapping categories; DANA crisis figures reflect repository classifications across 83 total fact-checks.

In 2026, several events have triggered major waves of disinformation. Perhaps the two most significant have been those associated with the migration crisis in Ceuta (EDMO, 3 September 2026) and the fatal high-speed train accident in Urda, in the province of Toledo, which occurred on 28 October, one month before the General Elections. At the time of publication of this report, just a few days after the accident, it is still too early to assess how this recent event may affect disinformation dynamics during the election campaign.

Other recurrent disinformation patterns

Migration narratives remain a constant focus across the repository: between 2017 and 2026, researchers logged 883 fact-checks addressing migration and racism, with 75% receiving a verdict of deliberate fabrication (IBERIFIER, 2026, p. 119). Migration represented the second-largest fact-checking category in 2025, accounting for 11.4% of all entries that year (Fig. 31).

Repository data shows domestic actors generate the vast majority of disinformation, with foreign accounts offering selective amplification. A recent IBERIFIER report (2026a, pp. 116–119) illustrates this dynamic across three cases: gender-related disinformation (368 fact-checks, driven primarily by domestic polarization with secondary foreign amplification); the April 28, 2025 power outage (which triggered 21 distinct conspiracy theories due to a 49-day delay in the official report, driven by domestic channels); and the V-16 emergency beacon mandate, where researchers found no foreign involvement.

IBERIFIER operational mechanisms and intelligence

Heading into the pre-electoral period, the IBERIFIER observatory is deploying its monitoring frameworks across digital media in Spain. The hub integrates the international TTP (Tactics, Techniques, and Procedures) taxonomy into its verification architecture to classify manipulation techniques—including narrative laundering, synthetic media, cross-platform amplification, and FIMI patterns. IBERIFIER also participates in the EU Code of Practice’s Rapid Response System (RRS), established under Measure 37.2 pursuant to Article 45 of the DSA. The RRS provides a direct alert mechanism connecting independent researchers, fact-checkers, EDMO hubs, and signatory platforms (VLOPs/VLOSEs), activating during election periods—from four weeks before the vote through one week after—and during crisis events.

During the elections, IBERIFIER activates ‘Comprobado’, A collaborative network to share monitoring workloads by fact-checkers

The observatory does not issue alerts without prior verification against primary sources, technical attribution tools, and cross-organizational review. When an alert shows signs of manipulation but lacks technical confirmation of falsity or deceptive intent, teams do not escalate it to the RRS.

During national elections, IBERIFIER will activate the ‘Comprobado’ collaborative network to share monitoring workloads by fact-checkers and cross-verify reporting across newsrooms. The observatory targets a response time under 24 hours from verified detection to alert issuance during crises, aiming for under 12 hours during peak election windows.

To protect observatory independence and uphold freedom of expression, IBERIFIER exercises no regulatory or enforcement authority. RRS submissions represent independent technical evaluations; the hub does not issue takedown requests or account suspension orders, nor does the RRS legally obligate platforms to take action. Moderation, labeling, and reach-reduction decisions rest solely with participating platforms under their service terms and DSA obligations. Neither the European Commission nor national governments flag content through the RRS; the Commission acts strictly as a logistical coordinator. The RRS operates alongside standard user reporting channels without replacing them.

Sources

• EDMO (September 3, 2026). Ceuta and the disinformation and foreign interference campaigns (FIMI): what we know and what we don’t. https://edmo.eu/publications/ceuta-and-the-disinformation-and-foreign-interference-campaigns-fimi-what-we-know-and-what-we-dont/
• Fillon, M., Kontosakou, A., McCluskey, C., Scott, G., Taylor, J., Ndep, J., & Thurbon, R. (2026, August 3). Not delayed, not deferred: EU AI Act transparency obligations are now in force. Goodwin.https://goodwinlaw.com/en/insights/publications/2026/08/alerts-technology-dpc-eu-ai-act-transparency-obligations-now-in-force
• IBERIFIER (Badillo, Á., Paisana, M., Arteaga, F., Izquierdo, D., Salaverría, R., Sánchez-Hualde, A., Cardoso, G., & Margato, A.) (2026a). Social impact of disinformation campaigns in Spain and Portugal 2024-2026. IBERIFIER. https://doi.org/10.5281/zenodo.20556457. 
• IBERIFIER (Rosso, P., Aloy Mayo, M., Huertas-Tato, J., Camacho, D., Rosas, C., Rementeria Nuñez, M. J., Salaverría, R., Garde-Eransus, E., & Sánchez-Hualde, A.) (2026b). Impact of AI in Spanish and Portuguese language disinformation. IBERIFIER. https://doi.org/10.5281/zenodo.21703744
• Iyer, R. (2025, July 25). Meta to stop selling political ads in the EU from October. TechCrunch. https://techcrunch.com/2025/07/25/meta-to-stop-selling-political-ads-in-the-eu-from-october 
• Larraz, I., & Salaverría, R. (2026). De los deepfakes al LLM Grooming: taxonomía de los nuevos usos de la inteligencia artificial para desinformar. In D. García-Marín (Eds.), El algoritmo de la mentira. Convergencias entre IA y desinformación (pp. 57-73). Tirant Lo Blanc.
• Palau-Sampio, D. (2023). Pseudo-media disinformation patterns: Polarised discourse, clickbait and twisted journalistic mimicry. Journalism Practice, 17(10), 2140-2158. https://doi.org/10.1080/17512786.2022.2126992
• Salaverría, R. (2026). Iniciativas contra la desinformación en español. Archiletras Científica, 16, 215–225. https://doi.org/10.5281/zenodo.23760680
• Salaverría, R., Rodríguez-Virgili, J., & Sierra, A. (2026). Disinformation and Elections. In Ginsborg, L., & Gori, P. (Eds.), Disinformation. A Multi-Disciplinary Analysis (pp. 213-226). Springer. https://doi.org/10.1007/978-3-032-00480-2_12
• Sánchez Torres, M., Magallón Rosa, R., & Pastor Marín, F. M. (2025). Desinformación y catástrofes naturales. El caso de la Dana de Valencia en 2024 : Análisis de los bulos difundidos a partir de las publicaciones de cuatro verificadores españoles. AdComunica, (30). http://dx.doi.org/10.6035/adcomunica.9015
• Universidad de Navarra (2026, 16 June). Digital News Report España 2026: interest in news grows and media keep credibility far above social media and AI. Press release, consulted online (https://www.unav.edu/noticias/-/contents/16/06/2026/el-interes-por-las-noticias-crece-y-los-medios-mantienen-una-credibilidad-muy-superior-a-la-de-redes-sociales-e-inteligencia-artificial/content/lovPblW1fC70/198397782); news avoidance from the UNAV Digital News Report blog.
  
• Other secondary sources:
  Third-party statistics and figures (Digital News Report, Evercom, GfK DAM, Maldita.es, Democracy Reporting International, EEAS, DSN, among others) are cited from IBERIFIER (2026) as secondary sources.

Cite this report as:

Sierra, A., & Salaverría, R., (2026). Preliminary Report ahead of 2026 General Elections in Spain. IBERIFIER. https://doi.org/10.5281/zenodo.23843925