COVID-19 Disinformation Report - Volume 2

Democrat Panic, media delegitimization and more bioweapons – Blackbird.AI analyzed growing disinformation narratives around the COVID-19 pandemic in early 2020. Part 2 of an ongoing series.

March 17, 2020




The COVID-19 (Coronavirus) outbreak is an evolving human tragedy affecting millions of people worldwide. In today’s digital media landscape, any major event, especially one like this, results in opportunistic disinformation dynamics around trending news that seek to cause harm towards brands, governments and the public. The COVID-19 outbreak is an unprecedented global event, in size and scope, resulting in constant trending events across social media at a faster pace than ever before. The head of the World Health Organization (WHO) calls the spread of false information on COVID-19 an “infodemic” resulting in widespread confusion around what we can and cannot believe, in the midst of an unprecedented, rapidly-shifting low information public health crisis.

Modern disinformation campaigns primarily seek to modify perceptions and beliefs that can influence the public, policy makers, government officials, and even nation-states. A solid understanding of the scope and scale of disinformation pertaining to COVID-19 is critical to every aspect of lives, both personal and professional, as these attacks will continue as long as the pandemic continues, with information-based decisions impacting communities around the world.


Blackbird.AI rapidly identified, through AI-driven insights, significant amounts of manipulated content across social media. We also detected anomalies in techniques being used to shift public perception through the manipulation of narratives. Our system revealed a variety of actors that introduced unnecessary uncertainty into the global information ecosystem. These actors included people, outlets, and entire networks of coordinated, inorganic amplifiers and bot networks.

Blackbird.AI’s COVID-19 Disinformation Report analyzed 49,755,722 tweets from 13,203,289 unique  users on topics related to COVID-19 for the period of February 27 – March 12, 2020*. We found 18,880,396 million tweets to be manipulated content, with a Blackbird Manipulation Index (BBMI) of 37.95%. The red curve shows inorganically amplified manipulative conversations and the region between the curves is a measure of the organic contribution.

The confluence of developments in Italy with a significant drop in the Dow Jones triggered a 300% increase in total Tweet volume as well as manipulated contribution to that conversation. As panic escalated the chatter, we noted significant piggybacking of other disinformation narratives to ride that wave.

*Our process of discovery uncovers new keywords as narratives unfold so the final volume statistics have been augmented with additional targeted searches. We’ve taken care to use a background set of data in the daily graphs for making a fair volume comparison day by day.

We have observed a distinct shift from the behaviors observed in our previous COVID-19 Disinformation Report, published on February 19, 2020. In the prior report, the disinformation flagged by our system was highly organized and synthetic with widespread use of bots pushing specific hashtag-driven narratives that were both targeted and consistent, which is typical of manipulation campaigns.

Our current findings exhibit organized, synthetic (bot-driven) disinformation, but this was far
exceeded by a massive amount of decentralized misinformation that was overwhelmingly made up
of individuals or groups “piggy-backing” onto more general hashtags related to COVID-19. It appears disinformation actors are currently shifting to seeding networks with deception, driving many users to spread these unknowingly. This insight is supported by a very high BBMI across a smaller number of general hashtags, e.g. #coronavirus, #COVID19 with predictably huge volume as demonstrated in the graph below.

These three hashtags related to the Coronavirus outbreak are high volume and exhibit high BBMI. The reason for this is that they are serving as vehicles to amplify disinformation narratives for unrelated topics through “piggybacking”. Some specific examples are included as part of the narrative expositions below.

In the following narrative discoveries, we offer a high-level forensics landscape analysis of the actors and networks contained in four of the largest system-surfaced disinformation narratives that occurred within our recorded time frame related to COVID-19.


Narrative 1: Dem Panic
Campaigns delegitimizing democratic leadership were enacted in conjunction with medical disinformation attempting to downplay early efforts at containing COVID-19.

Narrative 2: Media Delegitimization
These campaigns targeted the mainstream media as purveyors of overblown concern and fear mongering. “Fake news” was one of the most manipulated terms deployed to delegitimize any official, media-sanctioned narrative regarding the pandemic and concerns around it.

Narrative 3: Bioweapons
As nations around the globe reckoned with COVID-19 within their own borders, several conspiracies—new and revived ones—floated the idea that the virus was man made.

Narrative 4: Immigration Xenophobia in Europe
As geopolitical tensions involving COVID-19 escalated across Europe, ethno-nationalist movements utilized the confusion to push their agendas.

Blackbird.AI uses machine learning, network analysis and propagation patterns to identify and combat disinformation. Our proprietary AI-Driven SaaS platform provides a sophisticated understanding of both manipulation and harmful digital media. The Blackbird Platform surfaces synthetic amplification from bots and bot-nets, delegitimization campaigns, and meme-driven propaganda.


We have quantified the total volume of manipulative tweets, which is a prototypical indicator of a disinformation campaign.

Manipulation leads to harmful content, defined as any content that can be used to drive harmful narratives

Disinformation is defined as false information intended to mislead; however, in today’s culture it has become a catch-all term for any manipulative or harmful information-spreading activities. Synthetically amplified manipulation campaigns blend fact and fiction to intentionally create a deleterious effect on their subject.


The Blackbird Manipulation Index (BBMI) is a proprietary system designed to measure the degree to which the content of a tweet exhibits traits of being manipulated, or a user appears to be engaged in manipulation.

The assessment is based on features we use to score content which blends
together attributes extracted from the data such as toxicity of language,
communication patterns, post volumes and bot activity.


Blackbird’s system screens large collections of posts using AI algorithms to look for subtle patterns in the data that suggest deliberate manipulation.

Our methods find:

  • Suspicious patterns of propagation and sharing
  • Manipulated messaging around specific content including hashtags, URLs, and images specific users and communities that are engaged in manipulation
  • Evidence of falsification that was introduced via technical means
  • Traffic that is generated via bots or coordinated user campaigns

Because the signals in the data that serve as ” tells” for these effects are spread throughout the data as it evolves over time, they are not easy to detect and analyze without sophisticated automation support.”

In Blackbird’s system, the content associated with high disinformation risk is automatically surfaced so that human intelligence can be brought to bear selectively and on the items that appear most relevant for interpretation. This capability complements and greatly enhances both the efficiency and effectiveness of analyst-driven search and exploration as they understand emerging situations in the data.

The understanding of disinformation factors is critical to ensuring the integrity of the data on which decisions of consequence are based.


Total Tweets: 2,535,059
Total Inorganic Tweets: 839,764
Total Users: 911,533
Total Languages: 50
Avg Tweets Per User: 2.78 times


Blackbird’s system uncovered significant manipulative narratives aimed at the Democratic Party and its leadership. The goal of this campaign appeared to be the de-legitimization of Democratic politicians and party members, framing them as overblowing or otherwise overhyping the dangers of COVID-19.

The driving message: that the Democrats were overblowing the issue in order to hurt President Trump. The “Dem Panic” narrative and related spin offs also included the widespread mention of the “out of control” homeless population and high number of immigrants in Democraticdistricts. Many of these messages unwittingly found their way into what would traditionally be considered credible media stories.

While the weaponization of disinformation for political purposes has been fairly common for several years now, this campaign spreading during a health crisis could be particularly harmful to the public; it could potentially obfuscate the real dangers of the COVID-19 pandemic. In particular, ignoring the recommendations that underpin the seriousness of COVID-19 can have significant detrimental impacts on communities across the country and the world.

An interesting component to this narrative is how intertwined online disinformation activity was with two particular events:

February 28 – President Trump’s Charleston, SC rally, during which he criticized the Democrats of politicizing the Coronavirus and characterizing his handling of the outbreak as “their new hoax” (referring to Democrats). The generic #coronavirus and other similar hashtags had been at the top of the list of most used hashtags throughout this national emergency, but following this statement other hashtags began to rise to the top, including: #hoax, #Democrats, #DemHoax, #FakeNews, #TrumpRallyCharleston and #MAGA.

March 9 – After the Italian government quarantined the entire country. In response to this, there was a significant spike in manipulated conversations in many of the same narratives that spiked following President Trump’s Charleston rally. It is likely this was a response to the uptick in posts expressing heightened concern about the virus and possible action needing to be taken within the United States. The increase in posts calling for immediate action regarding the virus was met with a paralleled increase in posts accusing Democrats of overhyping the issue for political reasons.


Critical Manipulated Entities

Manipulated Hashtags

  • #QAnon– 2300 posts – 63.38% BBMI
  • #homeless – 6431 posts – 60.83% BBMI
  • #Homeless – 3111 posts – 55.41% BBMI
  • #ShahidVsPelosi – 2249 posts – 60.10% BBMI
  • #KAG – 6019 posts – 58.94% BBMI
  • #MAGA– 9801 posts – 57.00% BBMI
  • #Pelosi – 6126 posts – 53.17% BBMI

Users of Interest

  • @EcoInternetDrGB (Suspended) – BBMI 97.2
    (high volume)
  • @Jim43520096 – BBMI 100
  • @MomJar1 – BBMI 86.5
  • @tooranomaki6910 – BBMI 100 (high volume)

At an evening rally on Feb 28th, President Trump criticized the Democrats for politicizing the Coronavirus and characterized their attacks on him for his handling of the outbreak as “their new hoax” (referring to Democrats). This led to a large spike in activity on this narrative on Feb 29th. Additionally, Italy announcing a national quarantine on March 9 led to an explosion of generic Coronavirus-related hashtags. As panic escalated the chatter, we noted significant piggybacking of other disinformation narratives to ride the wave on these common hashtags.



Total Tweets: 309,571
Total Inorganic Tweets: 130,107
Total Users: 193,341
Total Languages: 42
Avg Tweets Per User: 1.60 times


Blackbird’s system uncovered manipulation campaigns designed to delegitimize the media, suggesting all of the mainstream media was “fake news” and that journalists were actively trying to create panic among the masses by sensationalizing the dangers posed by the COVID-19 pandemic.

These manipulation campaigns included widespread accusations that the Democratic Party controlled the media, and was leveraging it to criticize the government response to the COVID-19 pandemic.

Out of all of the manipulation campaigns highlighted in this report, the attempts at media delegitimization are likely the most dangerous to public health. Casting doubt on the media can potentially lead to disbelief in what is being reported, often including warnings and directions from government officials, public health agencies, and experts in the field. This could lead to certain pockets of the population not listening to these requests and endangering themselves and their communities during the COVID-19 pandemic.


Critical Manipulated Entities

Manipulated Hashtags

  • #CNN – 1,342 posts – 60.18% BBMI
  • #MSM – 5,367 posts – 47.97% BBMI
  • #Hoax – 1,692 posts – 47.38% BBMI
  • #FakeNews – 18,197 posts – 46.72% BBMI
  • #fakenews – 3,310 posts – 44.55% BBMI

Users of Interest

  • @BRNordgren – 94.0 BBMI
  • @ChristinePolon1 – 75.3 BBMI
  • @LivePDDave1 – 71.2 BBMI
  • @Jali_Cat – 54.27 BBMI

There is understandably some overlap in the activity between the “Dem Panic” and “Media Delegitimization” narratives, and for that reason a similar pattern emerges. There is a smaller spike this time following President Trump’s Charleston rally referencing “their new hoax”, and the familiar explosion (both organic and manipulated) after Italy announces their nationwide lockdown caused by piggybacking on generic Coronavirus-related hashtags.



Total Tweets: 1,299,488
Total Inorganic Tweets: 553,437
Total Users: 770,720
Total Languages: 60
Avg Tweets Per User: 1.69 times


Blackbird’s system uncovered broad manipulation campaigns suggesting the COVID-19 pandemic was intentionally created by the “global elite”. Themes revolved around “population control”, profit from “the cure”, state actors creating the virus as a biological weapon. While these narratives differed slightly with some saying the virus was released accidentally while others suggested it was done intentionally, they all revolve around the current pandemic being the handiwork of a group or organization.

Several manipulation campaigns suggest that, in their attempt to develop a biological weapon and expand biological warfare capabilities, China, the United States, or Iran was responsible for the creation and release of the virus. These claims were often accompanied by fabricated scientific documentation presented as evidence to support the accusation. The idea that there will be intentional depopulation and societal collapse as a result of these conspiracies were pervasive within these campaigns as well.

This campaign and its associated narratives mirrored one of the main themes outlined in this report -that disinformation was initially created synthetically and amplified (often by bots) before being both consumed and amplified more organically by individuals in a decentralized way. It is likely that those responsible for the synthetic creation of this campaign understood society’s collective fascination with conspiracy theories and more importantly, understood that a large number of social media users would believe these stories and amplify them.


Critical Manipulated Entities

Manipulated Hashtags

  • #WuhanVirus – 24,350 posts – 50.56% BBMI
  • #BillGates – 4,475 posts – 50.30% BBMI
  • #AyatollahsSpreadCOVID19 – 2,954 – 74.93% BBMI
  • #wuhanvirus – 1,932 posts – 98.55% BBMI
  • #bioweapon – 4,170 posts – 70.72% BBMI
  • #ChinaCoronaVirus – 1,477 posts – 59.87% BBMI
  • #BioWeapon – 1,517 posts – 57.98% BBMI
  • #WuhanPneumonia – 1,230 posts – 56.44% BBMI
  • #WuhanCoronavirus – 3,365 posts – 56.02% BBMI

Users of Interest

  • @MaggiePeggy123 – 96.8 BBMI
  • @ZyroFoxtrot – 87.8 BBMI
  • @tooranomaki6910 (again) – 99.7 BBMI
  • @Miladsaboori3 (Iran) – 96.1 BBMI
  • @Free_HK_Uighurs (Suspended) – 90.8 BBMI
  • @BlancheBullshit – 98.3 BBMI
  • @DIXIEDOODLE (Suspended) – 93.19 BBMI
  • @halimahbdavid (Suspended) – 100 BBMI
  • @RedState – 86.4 BBMI

In this graph again the same spike occurs around March 9th, driven by the full-country lockdown in Italy and users piggybacking on common hashtags to spread high volumes of both organic and manipulated content. An interesting additional insight is that the toxic hashtag #AyatollahSpreadsCOVID19 spiked on March 7th, creating another mini-spike starting on that day.



Total Tweets: 29,731
Total Inorganic Tweets: 16,250
Total Users: 22,638
Total Languages: 31
Avg Tweets Per User: 1.3 times


Blackbird’s system identified disinformation campaigns portraying immigrants crossing into Europe as likely carriers of COVID-19. As part of their campaign and associated narratives, European white nationalists labelled migrants crossing into Southern Europe from Turkey as “a plague”. Beyond geopolitical tensions involving COVID-19, European ethno-nationalist movements utilized the confusion over the facts surrounding the virus and its spread and fear of the unknown.


Critical Manipulated Entities

Manipulated Hashtags

  • #IStandWithGreece – 7,774 posts – 49.55 BBMI
  • #GreeceUnderAttack – 7,517 posts – 47.26BBMI
  • #Turks – 2,795 posts – 43.05 BBMI
  • #Erdogan – 2,687 posts – 44.71 BBMI
  • #GreeceDefendsEurope – 1,067 posts – 46.89 BBMI

Users of Interest

  • @BasedPoland – 90.0 BBMI

This graph shows a spike in manipulated content around March 7th, following the lockdown in Northern Italy.

We're hiring, See our careers page. Learn More