The regime of Nicolás Maduro in Venezuela has been spreading disinformation with artificial intelligence- (AI) generated newscasters, who only deliver news favorable to the regime, highlighting how the technology is used to further its narrative, Spanish daily El País reported.
According to the report, based on a study by Venezuelan nongovernmental organization (NGO) Cazadores de Fake News, both ‘journalists’ Daren and Noah, who appear in various fictitious digital newscasts speaking English, such as the Venezuelan newscast House of News, were created with AI from the Synthesia software’s catalog of more than 100 multiracial faces.
“Regimes like Venezuela or China create fake avatars, photographs, and videos with a political purpose, distorting reality,” Victor Ruiz, founder of the Mexico-based cybersecurity center SILIKN, told Diálogo on April 10. “This way they retain power over people by making them distrustful of people, institutions, or even nations that might represent competition to the regime.”
For about $30 a month, Synthesia users can input a script that can be generated in more than 100 languages and made to sync with the mouth of avatars and be included in images, soundtracks, videos as part of the editing, Spanish news site Prensa Ibérica reported. No knowledge of video creation is required for its operation.
According to Synthesia’s terms of services, its website can be used to create training, tutorial, and marketing videos; while religious, political, sexual, discriminatory, hate speech, or blasphemous content is restricted. In addition, the London-based company created in 2017 stresses that stock avatars are not to be “in user generated content for TV broadcasting.”
Héctor Mazarri, collaborator of Cazadores de Fake News, told El País that after studying Daren and Noah’s videos they detected “an organized attempt to push narratives favorable to the [Venezuelan] regime. While a trained user can detect the errors, this is done so that no one is exempt from not believing it,” he said.
Furthermore, the videos of the ‘journalists,’ unilaterally favoring Venezuela without basis, had hundreds of thousands of views on YouTube, went viral on TikTok, and were broadcast by state-owned Venezolana de Televisión, El País reported. “We can all be victims of this abuse of power and disinformation,” Ruiz added.
In a House of News broadcast, one of the avatars shared a positive view of Caracas. “We wanted to know if Venezuela is really as destroyed as the media has claimed for years,” he said, according to El País, before showing images of Venezuelan beaches full of tourists and announcing that travel agencies have been fully booked for the carnival season.
The thrust of this video was to create the perception that economic and social reports of Venezuelan institutions and civil organizations, which indicate that more than 90 percent of the population is living under the poverty line, aren’t accurate, Venezuelan news site Efecto Cocuyo reported.
“Countries like Venezuela and China use AI and other new tools as weapons against people, against the citizenry, or another country,” Ruiz said. “This makes the people distrust, have uncertainty, and potentially rebel against the country itself.”
Beijing, for its part, has also been promoting the interests of the Chinese Communist Party (CCP) through an AI-generated avatar in a newscast called Wolf News. In one video, for instance, a fake anchorwoman heralded China’s role in geopolitical relations at an international summit, The New York Times reported on February 9.
Graphika, a U.S. firm that studies disinformation, located a campaign promoting pro-China avatar video footage in 2022. “The use of publicly available AI products will enable influence operation actors to create misleading content at greater scale and speed,” it said.
“It’s a complex issue, in the case of China there are barriers that prevent first-hand knowledge of what’s really going on,” Ruiz said. “Many more avatars are going to be created in authoritarian countries with political motives. We have to be very cautious with the information we receive.”
In March, the CPP’s official media outlet People’s Daily unveiled an AI news anchor. The virtual journalist named Ren Xiaorong can only respond from a dictated script that follows the CCP’s editorial and official line, Argentine news site TN-Todo Noticias reported.
Codes of ethics
Given that AI is here to stay, “organizations and governments must have codes of ethics for their operations. They have to work immediately to incorporate regulations for the use of AI,” Ruiz said. “These codes could somehow curb some of the misinformation.”