Pre-election 2026 parliamentary elections campaign monitoring: additional insights from Facebook

Executive summary

This report conducted under the Bulgarian–Romanian Digital Media Observatory (BROD) covers the second weekly monitoring cycle of Facebook public content during the Bulgarian pre-election campaign. Its purpose is descriptive and exploratory: to characterise the volume, distribution, actor composition, and narrative patterns of publicly accessible Facebook content in a defined pre-election window, using computational methods applied to a structured corpus. No causal claims about influence, no sentiment judgements, and no claims of representativeness beyond the observation window are made.

A keyword search for „избори“ ("elections" in Bulgarian) via Meta Content Library returned more than 7 000 posts from public pages, groups, and profiles across the seven-day window (March 30 – April 5, 2026). Of them, less than a third remained in the downloadable subset that includes posts from pages with 15K+ likes or followers, as well as posts from profiles that are verified or have 25K+ followers.

Option 1 selected, downloadable subset: 1,300 estimated total results

All posts were publicly accessible at the time of collection; no private or group-restricted content was included.

The narrative monitoring framework was drawn from an existing media and social media listening document provided by the research team. It comprises 11 narrative clusters, 23 sub-clusters, and 181 keyword and phrase search terms. Keyword matching is purely lexical; no supervised classifier was applied.

The analysis was organised into the following completed stages:

Stage 1 – Data cleaning and classification.

Stage 2 – Engagement statistics and network analysis: descriptive statistics (non-parametric), reaction-type breakdown, temporal patterns, and top-actor ranking by engagement. Two network graphs were constructed: an owner-to-surface bipartite network (51 nodes, 54 edges), and a co-text coordination network (100 nodes, 141 edges, 35 Louvain communities).

Stage 3 – Per-actor analysis across all 23 actors (that actually have measurable presence in the dataset -- not the full register count): mention counts derived from text regex on both the owner axis and the mention axis, per-actor engagement metrics, and reaction-type profiles.

Stage 4 – Narrative keyword analysis: lexical matching against the 181-term framework, producing match counts by narrative cluster and by actor.

Stage 5 – Bulgarian NLP pipeline using stanza 1.11.1 (Bulgarian model): full corpus lemmatisation, part-of-speech tagging, and named entity recognition across 140,608 tokens and 9,444 unique lemmas after stopword filtering. Outputs include corpus-wide lemma frequency tables, per-actor lemma profiles on both axes, log-likelihood distinctive vocabulary per actor, collocate analysis in a ±5-token window for 18 narrative keyword stems, KWIC concordance (472 lines), hashtag frequency and co-occurrence analysis, and URL and domain extraction from all four URL-bearing columns (514 URL instances across 453 posts, 137 unique domains).

Key methodological decisions

Several decisions warrant explicit statement.
No sentiment claims are made anywhere in the analysis.

Exact-text duplicates were retained rather than deduplicated, because coordinated identical posting is treated as a substantive phenomenon rather than noise. Posts attributed to ИТН, for example, are 97 % reshares with 58 exact-text duplicates; removing them would suppress the coordination signal entirely.

Named entity recognition from the stanza Bulgarian model was retracted after post-hoc verification revealed systematic unreliability for entity counting in this corpus. Entity frequencies reported in the analysis derive from direct regex-based counts, not from the NER output. This substitution is conservative: regex counts are transparent and reproducible, though they do not resolve ambiguous entity references.

Keyword matching for narrative classification establishes that a narrative term is present, not that a post endorses, promotes, or opposes the associated frame. Directionality was partially resolved for a subset of terms through Stage 5 collocate analysis. Collocates of "суверенитет" (защита, защитавам, европейски, лидер), for instance, indicate predominantly pro-EU defensive usage, which partially resolves the directionality problem for narrative cluster N1. Where collocate evidence is insufficient, directionality remains unresolved and is reported as such.

The mention axis and the owner axis are treated as analytically distinct throughout. A post may be owned by one actor while mentioning several others; conflating these axes inflates mention counts and misattributes engagement.

Limitations

The observation window of one week is short relative to the full campaign cycle. Patterns identified here may not hold across the broader pre-election period, and no such generalisation is attempted.

The corpus is bounded by what Meta Content Library returned for the specified query parameters and by the downloadable subset restrictions which can be quite limiting and hinder valuable insights into a low-resource language. Retrieval completeness cannot be independently verified; systematic omissions, if present, are not detectable from the dataset alone.

Keyword matching carries well-understood precision and recall trade-offs. The 181-term narratives framework was designed for broad coverage; false positives are expected at some rate and cannot be quantified without manual annotation, which has not been performed.

The 200 posts matched via OCR image-text extraction constitute a lower-reliability subset. Error rates for Bulgarian Cyrillic text in naturalistic image conditions are not characterised for this specific pipeline; findings derived primarily from this subset should be treated with corresponding caution.

No human annotation or manual content coding has been conducted, and computational outputs have not been validated against a gold-standard labelled sample. The current analysis is better described as a structured quantitative description than as a full mixed-methods content analysis.

Finally, engagement metrics reflect platform-mediated amplification dynamics and not audience attitudes or reach in any direct sense. Reaction counts measure a behavioural signal whose interpretation requires caution, particularly given the extreme concentration of attention (Gini = 0.856) across a small number of posts and actors.

Donwload the report [+]

At the beginning of 2026: Prebunking narratives in Bulgaria

In 2026, Bulgaria will hold a series of elections - early parliamentary elections and regular presidential elections. The dates are not yet known, but from now on, we can say what narratives may emerge, and this initial list can serve as a step towards disarming them or, in other words, as a prebunking tool.

The narratives that we could expect to emerge during the year are often well-known. Here we use previous fact-checks and narrative reports, made mainly by EDMO and its hubs, and by other approved fact-checking organizations. [+]

Inside the Facebook Feed Before Bulgaria’s 2026 Election: Narrative and Manipulation Pattern Mapping:  March 30 – April 5, 2026

This is the second edition of Narrative Atlas – a weekly analytical update of narratives within the Bulgarian information environment on Facebook - which focuses on the upcoming Bulgarian Parliamentary Elections, the eight rounds of national elections since 2021. [+]

Defending the Vote: Narrative Pressure and Infrastructure Risks Update

With Bulgaria's parliamentary elections on 19 April 2026 approaching, this briefing documents the active narrative pressure and infrastructure risks identified throughout February and March 2026.

The briefing maps the operational behaviour of Bulgaria's domestic disinformation amplification ecosystem, tracking how pro-Kremlin content is flooded, laundered from fringe to mainstream discourse, and strategically amplified across platforms. Key findings include the rapid activation of the Pravda network around the seizure of Ukrainian assets in Hungary, the systemic republication of content from EU-sanctioned Russian entities, and the demonstrated effectiveness of Telegram as a high-reach distribution layer with 51% of Pogled Info's reposted videos accounting for nearly 88% of its total monthly reach. The briefing also examines the "Petrohan" case as a live example of narrative laundering from fringe outlets into parliamentary discourse.

Structural vulnerabilities identified include content-flooding capacity, sanctions enforcement gaps, and the under-monitoring of TikTok and short-form video platforms. The briefing is designed as an executive risk update and does not assess the accuracy of political statements, focusing instead on how narratives move and gain traction through documented amplification infrastructure. [+]

Inside the Facebook Feed Before Bulgaria’s 2026 Election: Narrative and Manipulation Pattern Mapping  March 23 – March 29, 2026

This is the first edition of Narrative Atlas – a weekly analytical update of narratives within the Bulgarian information environment on Facebook - which focuses on the upcoming Bulgarian Parliamentary Elections, the eight round of national elections since 2021. Bulgaria’s political and informational environments continue to be characterised by widespread distrust and polarisation[1]. By offering these findings, it is our hope to contribute to a more holistic understanding of how election campaign narratives function and position themselves. In addition, by cataloguing and analysing these patterns, we believe we can aid in the larger work of slowly building societal resilience to false, misleading or manipulative information through education and online campaigns. [+]

Defending the Vote: Policy Responses to Information Warfare in Bulgaria

As Bulgaria heads into its ninth general election since 2021 the country faces sustained Russian information manipulation pressure compounded by a deeply embedded domestic amplification ecosystem.

This report provides an anticipatory threat assessment for the April and November 2026 votes, drawing on CSD’s extensive Kremlin Playbook research and methodology, network and algorithmic analysis, and narrative monitoring. It maps the architecture of manipulation operating in the Bulgarian information space, identifies the irregular warfare narrative pressure points most likely to be exploited during the campaign, and outlines actionable recommendations – grounded in the European Digital Services Act Elections Toolkit and comparative European experience – for closing the gap between awareness and operational preparedness. [+]

MVI Methodology: Monitoring Framework for Code of Conduct on Disinformation

Research partners in Bulgaria and Romania and the European Observatory of Digital Media have created a first of its kind methodology (Materiality, Verifiability, Impact, or MVI) for monitoring the compliance of Very Large Online Platforms and Search Engines with the EU’s strengthened Code of Conduct on Disinformation. This report is situated within the evolving regulatory landscape of the Digital Services Act (DSA), which, as of July 2025, elevated the Code from a voluntary commitment to a formal compliance benchmark. [+]

Public Attitudes and Security Perceptions in Bulgaria: Resilience and Vulnerabilities in 2025

The latest data from GLOBSEC Trends 2025 for Bulgaria. [+]

To Make Someone Do Something: Mining Alert-Style Directives in Bulgarian Social Media

A research, addressing a critical gap in the field of Natural Language Processing (NLP) for Bulgarian - a language often classified as "low-resource" due to the scarcity of annotated datasets, was presented at the LoResLM (Language Models for Low-Resource Languages) workshop during EACL 2026. [+]

A Multilingual, Large-Scale Study of the Interplay between LLM Safeguards, Personalisation, and Disinformation

Large Language Models (LLMs) can generate human-like disinformation, yet their ability to personalise such content across languages and demographics remains underexplored. This study presents the first large-scale, multilingual analysis of persona-targeted disinformation generation by LLMs. Employing a red teaming methodology, we prompt eight state-of-the-art LLMs with 324 false narratives and 150 demographic personas (combinations of country, generation, and political orientation) across four languages–English, Russian, Portuguese, and Hindi–resulting in AI-TRAITS, a comprehensive dataset of 1.6 million personalised disinformation texts. Results show that the use of even simple personalisation prompts significantly increases the likelihood of jailbreaks across all studied LLMs, up to 10 percentage points, and alters linguistic and rhetorical patterns that enhance narrative persuasiveness. Models such as Grok and GPT exhibited jailbreak rates and personalisation scores both exceeding 85%. These insights expose critical vulnerabilities in current state-of-the-art LLMs and offer a foundation for improving safety alignment and detection strategies in multilingual and cross-demographic contexts. [+]

The disinformation landscape in Bulgaria 2025

A new report about the disinformation landscape in Bulgaria [+]

Electoral Disinformation Ecosystems in Romania and its Diaspora: Cross-Platform Dynamics and Strategic Narratives during the 2024–2025 Electoral Cycle

This report investigates the disinformation landscape across X, TikTok, and Facebook during Romania's super 2024–2025 electoral cycle, with particular attention to the role played by Romanian diaspora communities, and a particular focus on those based in Italy. [+]

GLOBSEC Trends 2025: Ready for a New Era?

The Centre for Democracy & Resilience at GLOBSEC introduces the 10^th edition of its annual public opinion report, GLOBSEC Trends 2025. Based on polling conducted in nine Central and Eastern European countries – Bulgaria, Czechia, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, and Slovakia – this year’s edition offers a comprehensive picture of how societies in the region view global powers, security threats, democracy, and resilience. [+]

Public Attitudes in Romania: Staying in the West With Some Doubts

Romanian society is very supportive of the country’s place in the Western alliance while expressing some doubts about the policies of the EU and NATO. Support for Romania’s membership in these institutions fell during the COVID-years but recovered swiftly after Russia’s aggression against Ukraine.  [+]