“How true is it that Venezuela is such a poor country?” asks a blond presenter, in English, in a report on the Venezuelan crisis. But he is not a journalist, nor is he a human: he is an avatar generated by intelligence artificialincreasingly used for disinformation.
Subtitled pieces of a fictitious newscast, “House of News Español”, favorable to the ruling chavismo and questioned by verification services, appeared as advertising in social networks and unleashed controversy in this country.
This is not the only case.
Already another fake media, “Wolf News”, used this type of technology to also disseminate propaganda videos, which an investigation by the specialized company Graphika attributed in February to Chinese actors aligned with the State.
A manipulated video in which U.S. Pres, Joe Bidenwarned about a “threat” of an “extraterrestrial nature” that flooded social networks and messaging services in February.
And last year, a montage of Ukraine’s ruler Volodimir Zelenski snorting cocaine and a rap by Eminem against Mexico’s President Andres Manuel Lopez Obrador ran.
It was all done with AI, according to AFP digital verifications.
“More and more regions are starting to refer to ‘deep fakes’ and AI-generated audiovisuals as a threat,” Shirin Anlen, a media technology specialist at NGO Witness, tells AFP.
The avatars for the “House of News” videos come from the creation program via. intelligence artificial Synthesia. The company reported the account that disclosed them on YouTube, which removed those contents.
“If you pretend to be a journalist, you pretend to give news, you will be banned,” Synthesia spokeswoman Laura Morelli tells AFP, who, without mentioning the user, said he “violated” its “conditions and policies” and “was immediately identified and expelled.”
“House of News” anchors Noah, blond, and Daren, with a black complexion, are part of Synthesia’s catalog of 93 avatars. The same is true of “Wolf News” anchors Alex and Jason.
Synthesia was also used in January to create videos of alleged Americans supporting a coup in Burkina Faso.
Morelli, however, assures that this type of misuse of the platform is an exception.
These programs “can be used by anyone with minimal knowledge,” warns Eduardo Mosqueira, PhD in computer science and professor at the University of Coruña, in Spain. He assures that it is not difficult to identify if a video is fake but it is difficult to determine its origin. “If they have been minimally careful I suppose it is impossible.”
“I am a robot”
Venezuelan state television station VTV reproduced one of the videos of “House of News” that reported millionaire profits from the Caribbean Series, an international winter baseball tournament that returned to the country after nine years.
“These are numbers that this media outlet reports in the United States and a good part of Latin America”, said host Barry Cartaya.
There are no official figures.
Venezuelan President Nicolas Maduro bristles at any suggestion that production is linked to his government.
“It is not the intelligence artificialis the intelligence The president lashed out at the end of February. “I am a robot,” he continued ironically, imitating a metallic voice and miming mechanized movements.
One of the “Wolf News” videos criticizes U.S. inaction in the face of gun violence. In another, the importance of “great power cooperation” between China and the United States is stressed.
“This is the first time we’ve seen a state-aligned operation use video footage generated by intelligence artificial to create misleading political content,” Jack Stubbs, vice president of intelligence of Graphika, author of the study that linked the videos to networks seeking to promote the interests of the Chinese Communist Party.
Facebook parent company Meta dismantles networks of fake accounts, including one in Cuba, that created unreal identities “to post criticism of government opponents,” says Ben Nimmo, global leader of Intelligence on Signature Threats, in a videoconference with AFP.
“Some of these accounts use profile pictures that were likely generated using intelligence artificialthat is, photos of people who do not exist,” he says. “This is a tactic used more and more often.”
Mosqueira estimates that “it is becoming easier and easier to program bots on social networks that impersonate real people and launch the messages you want to get across to the public.”
The expert warns about the high impact of “deep fakes”, even when they are proven false.
“You will already know the famous phrase attributed to (Nazi propagandist Joseph) Goebbels that a lie adequately repeated 1,000 times becomes a truth.” (AFP)