Twitter released a dataset containing 1,196 accounts located in Venezuela that it said appeared to be engaged in a ‘state-backed influence campaign’ targeting Spanish-speaking audiences. The network operated during a period of social unrest in the country and often used automation tools in an attempt to influence public discourse on domestic issues as well as foreign affairs. Twitter content-management tools such as tuitutil.net (‘tweet’ in Spanish) and roundteam.co were among the most shared URLs in the network, which is again potentially indicative of coordination. DFRLab analysis also found that a large number of tweets were made using an application named RTNicolasMaduro that was linked to President Nicolás Maduro and his political party. Broadly, the network’s intent seemed to be shoring up support for the country’s socialist government (which has been led by Maduro since 2013, when he took office following the death of his predecessor, Hugo Chávez), as well as the country’s military. Twitter has also suspended accounts associated with Venezuelan government and military organisations. The accounts shared a range of images, including various pro-regime memes and video content targeting activists such as Lorent Saleh, who led protests against the Chávez government.
US-facing accounts
In January 2019, Twitter also removed 764 accounts located in Venezuela that it wasn’t able to tie to a government-linked information campaign. However, it suggested at the time that the accounts were ‘another example of a foreign campaign of spammy content focused on divisive political themes’. It also said that the behaviour was ‘similar to that utilized by potential Russian IRA [Internet Research Agency] accounts’. In June 2019, however, it added 33 additional accounts to the mostly English-language network, concluding that they were operated by a commercial entity originating in Venezuela.
While 97.08% of the 788 accounts were set to Spanish language (only 2.92% were in English), the account profile descriptions were overwhelmingly in English. They assumed a number of personas: Pro-Donald Trump (‘Boy mom! I love football! Conservative! I love my Country????! Small Business Owner! I work for a living! Catholic! I love God! Fitness Freak?? #MAGA #TRUMP’) and journalists covering the President (‘Analyst, journalist and writer, I dedicate myself to inform the people about all the controversy of United States, and the movements of Donald Trump, follow me’). A number of profile descriptions were also replicated across the network. For example, ‘Follow me, I am a journalist passionate about taking you all the secret information of President Donald Trump, #Trump #trumptop #journalist’ was used by six accounts.
