Executive Summary

The Modern Democratic Party of Moldova (PDMM), the successor to the „Pro-Moldova” formation and the heir to the political network built by fugitive oligarch Vladimir Plahotniuc, tried to re-enter the electoral competition under the leadership of Boris Foca and founder Vladimir Cebotari. The formation was, however, excluded from the parliamentary race by the Central Electoral Commission, on the grounds that the Public Services Agency had temporarily suspended it from the register of political parties, invoking „potential subversive actions”. In the public sphere, the decision was also associated with the migration of certain political figures from Ilan Șor’s entourage to PDMM, which amplified suspicions about the continuation of the influence mechanisms of oligarchic and pro-Kremlin networks in Moldova. PDMM is also perceived as an attempt to rebrand the old „party-state” from the years 2014-2019, appearing shortly after Plahotniuc’s public appeals to his former allies and intended to maintain his political influence.

Executive Summary

The Modern Democratic Party of Moldova (PDMM), the successor to the „Pro-Moldova” formation and the heir to the political network built by fugitive oligarch Vladimir Plahotniuc, tried to re-enter the electoral competition under the leadership of Boris Foca and founder Vladimir Cebotari. The formation was, however, excluded from the parliamentary race by the Central Electoral Commission, on the grounds that the Public Services Agency had temporarily suspended it from the register of political parties, invoking „potential subversive actions”. In the public sphere, the decision was also associated with the migration of certain political figures from Ilan Șor’s entourage to PDMM, which amplified suspicions about the continuation of the influence mechanisms of oligarchic and pro-Kremlin networks in Moldova. PDMM is also perceived as an attempt to rebrand the old „party-state” from the years 2014-2019, appearing shortly after Plahotniuc’s public appeals to his former allies and intended to maintain his political influence.

Why did we look at the Tik Tok page of the Modern Democratic Party of Moldova?

We analyzed the comments on the TikTok page of the Modern Democratic Party of Moldova (PDMM) and they indicate the existence of coordinated messages. The dominant aim is twofold: on the one hand, sprucing up PDMM’s image as a “democratic voice” repressed by the authorities; on the other hand, eroding public trust in the European path of the Republic of Moldova, in democratic processes, and in President Maia Sandu, through a succession of toxic messages and repackaged insinuations. The linguistic features of a substantial part of the comments betray generation assisted by artificial intelligence models. The formulations are sterilely correct and with an impersonal tone.

Thus, we discovered almost 300 comments supporting the PDMM party or blaming the PAS party and Maia Sandru, which were created by 59 fake accounts in Romanian, but also in Russian.

Besides the exact duplicates, the system discovered nine pairs of near-duplicate comments, that is, texts that are almost identical, with minor differences in wording. The most obvious are in Russian, with a degree of similarity of over 90%.

An eloquent example is the pair: „Когда была ПДМ, решения принимались обдуманно и профессионально. ПАС только разр...” and „Когда ПДМ был у власти, решения принимались продуманно и профессионально. ПАС то...”. The minimal lexical difference shows that the messages were intentionally rearranged, likely to avoid automatic detection, while preserving the same propagandistic idea.

This practice also highlights the existence of an astroturfing campaign: a strategy by which the illusion of popular support is artificially manufactured. Networks of fake accounts post bursts of copy-paste or near-duplicate messages, while real people pick up and multiply the pre-set texts to cover the traces of automation. The result is an apparently “green” surface of civic support, meant to convey the perception that PDMM would have a solid social base and a message that resonates in society, although in reality we are talking about artificial support built through coordinated digital manipulation.

The profiles fueling these waves have recurring hallmarks: generic names or first-name-plus-numbers combinations, stock avatars or the absence of authentic photos, vague bios, a lack of activity outside the targeted political topics, and a disproportionate preference for PDMM content. In addition, many of these profiles post at hours that are unnatural for the local audience (in the nighttime intervals 01:00–04:00).

Seen as a whole, everything looks more like a digital scheme that rests on three tactics:

• Using victimhood („we are excluded from elections and persecuted just because we are in opposition”);
• Delegitimizing the democratic institutions of the Republic of Moldova („the regime commits abuses, and the elections will not be free and fair”);
• Adopting the Kremlin’s propagandistic narratives („the EU and PAS threaten identity, neutrality protects us, the collective West brings only poverty”).

The propagandistic messages circulate in short formats – texts and clips – sometimes generated or rewritten with the help of AI. They are taken up and recycled from statements by pro-Kremlin Moldovan politicians or from their local proxy networks. The aim is not only to have PDMM’s image artificially inflated, but above all to create an appearance of discursive „normality,” in which the Europeanization of the Republic of Moldova appears as a threat, and pro-Russian radicals come to seem like reasonable options

Methodology

To identify coordinated inauthentic behavior (CIB) on PDMM’s TikTok page, we analyzed patterns of artificial amplification, the distribution of duplicate content, and mechanisms of automated engagement. The analysis focused on the comment ecosystem around PDMM’s official presence on TikTok, to determine the existence of coordinated networks that create artificial support.

Important specifications for data accuracy:

• The analysis focuses on the entirety of comments posted on PDMM content during the monitored period;
• Interaction indicators are analyzed within the limits of the access offered by the platform;
• Daily and hourly fluctuations in comment patterns;
• Statistical anomalies in the distribution and similarity of content;
• Assessment of the authenticity of the profiles that comment.

Data Collection and Processing:

Data collection: Comments were extracted with a TikTok scraper, capturing metadata (ID, text, username, timestamp, video ID, likes/replies).

Tools and Analytical Technique:

• Duplicate detection: Identification of identical comments through exact grouping
  Near-duplicate detection: 9 pairs of nearly identical comments (similarity score >0.8) using SequenceMatcher and Levenshtein algorithms;
• Temporal analysis: Examination of hourly distribution to detect bursts of activity;
• Profile evaluation: Analysis of generic usernames, lack of authentic images, empty bios;
• AI content detection: Identification of comments with a rigid style, grammatically perfect but unlikely for TikTok;
• Network analysis: Reconstruction of account networks based on co-occurrence in duplicates.

Methodology Limitations

• Access restrictions: TikTok limits access to complete data;
• Algorithmic variations: Frequent changes to the platform’s algorithms can influence patterns;
• Temporal context factors: Current events can influence engagement without involving artificial coordination.

These limitations were taken into account in the interpretation of the data, with manual validation processes to minimize false-positive classifications.