News

Venezuela Is Using Fake AI American Newscasters to Spread Disinformation

The same A.I. technology has also appeared to promote propaganda in Burkina Faso and China.
venezuela-ai-newscaster-disinformation
Noah is an A.I avatar posing as a journalist who features in YouTube videos that promote Venezuela. Photo: House of News Español via YouTube.

Strange videos promoting Venezuela are being shared on state run TV, using apparently English-speaking American hosts. 

The videos first appeared on a YouTube channel called the House of News Español, one featuring a blond white man speaking in American English over a clip called “Venezuela and its economy, myth or reality?”

Advertisement

But the purported journalist on screen is Noah, one of dozens of artificial intelligence avatars created by a company called Synthesia that for just $30 a month will say whatever the buyer wants. Similar videos produced recently using Synthesia software have been used to promote propaganda in Africa and Asia as well over the past month.

Noah is the host of a seemingly Western news broadcast that aims “to find out if Venezuela is really as destroyed as the media has claimed for years.”

Venezuela under President Nicolás Maduro and his socialist party has become one of the world’s most isolated countries and has long experienced chronic hyperinflation and food shortages. Allegations of state-run drug trafficking abound, as does a climate of silence after years of cracking down on the country’s free press. 

But according to Noah and House of News, all of that is overblown.

The nearly two minute long video—half of which is just video footage of a cell phone with someone speaking in Spanish about the various price options for all-inclusive packages to Venezuela’s tourist hotspots for carnival—has 270,000 views on YouTube, and is one of five odd videos posted on the channel within the last month. Another video with over a 100,000 views hosted by a different American sounding avatar details how Maduro’s political rival, Juan Guaido, allegedly spent $150 million. A third promotes Venezuela’s new world class baseball stadium.

Advertisement

What’s not clear is whether the Maduro administration was involved in the making of the videos. “The ecosystem of propaganda in Venezuela feeds on any type of content that could be useful to fuel the narratives of the Government of Nicolás Maduro,” Adrián González, the executive director of the Venezuela-based non-governmental organization Cazadores de Fake News (Fake News Hunters) who first brought attention to the videos, told VICE World News.

After being posted on YouTube, the videos began being shared on Twitter and Tik Tok, with minor pro-government personalities touting them as positive press about Venezuela from foreign news outlets. This became particularly problematic, González said, when the A.I. created videos then appeared on state-TV, with news hosts playing the videos directly from Tweets on their shows. This tactic places the A.I. created news reports on state TV while allowing the hosts to skirt responsibility for the content of the videos, because it came from social networks.

“The theory is that this is all a part of an organized strategy,” said González. “I cannot directly attribute these videos to the government of Nicolás Maduro precisely because this type of covert accounts and covert influence operations are designed so that there is no link that allows attribution.”

Although the videos are in English, they have Spanish headlines and bios on YouTube, as well as Spanish subtitles. The hosts speak with poor grammar and use odd words like “paradisiacal” to describe Venezuela’s beaches.

Advertisement

“There’s a lot of people outside of Venezuela that would say, well, I saw the video, but it has broken English and it doesn’t seem real. But Venezuelans don’t notice that really,” said González. “They are created for the Venezuelan public.”

The use of Synthesia’s technology was also recently used to prop up the military regime of Burkina Faso’s Interim President Ibrahim Traore that seized power during a coup in September.

In late-January, videos began spreading through WhatsApp groups and Facebook in Africa with American-sounding people supporting the government.

“Hello to the African people and particularly to the Burkinabe people. My name is Alisha and I’m a Pan-Africanist,” said one woman, mispronouncing the word “Africanist.”

“I appeal to the solidarity of the Burkinabe people, and the people of Burkina Faso to effectively support the authorities of the transition,” the A.I. avatar said. “Let us all remain mobilized by the Burkinabe people in this struggle. Homeland or death. We shall overcome.”

Synthesia said in February that the user who created the Burkina Faso videos was identified and banned for breaking a policy that allows the spreading of misinformation, disinformation or obscenity.

“As a company, we invested very early into content moderation to ensure our tens of thousands of customers benefit from a safe platform,” Synthesia CEO Victor Riparbelli told VICE World News in early February. “Cases like this highlight how difficult moderation is. The scripts in question do not use explicit language and require deep contextual understanding of the subject matter. No system will ever be perfect, but to avoid similar situations arising in future we will continue our work towards improving systems.”

Synthesia did not confirm whether or not the videos about Venezuela violated their content moderation guidelines when contacted by VICE World News this week.

Synthesia was founded in 2017 and uses video footage to clone actors, who are paid and give consent for the use of their likeness within the technology. The company recently told Motherboard that its primary clients use the technology to create real estate tours and HR training videos for corporations.

But Synthesia also appears to have quickly found another niche, whether wanted or not—controversial political figures and their supporters who want to spread misinformation.

Update: Shortly after publication, a Synthesia spokesperson told VICE World News that "The user in question has been identified and banned from our service due to a breach of our ToS. We're conducting an urgent review of our moderation process and will provide an update on the results in due course."