Media in several countries around the world have joined the artificial intelligence (AI) boom, either by adapting the use of Chat GPT to the publication of content or, in more risky instances, by creating avatars to replace television anchors.
China and Kuwait were the first to bring out virtual news anchors. However, Latin America has not been left behind.
In Mexico, in March, Grupo Fórmula introduced Nat, a presenter created by AI, who provides news summaries of Mexican and world events. In Venezuela, Nicolás Maduro's regime uses avatars to disseminate propaganda. And in Peru, the School of Arts and Human Sciences of the Universidad Nacional Mayor de San Marcos created a newscast starring a virtual anchor who speaks Quechua.
Journalists and activists are concerned that this technology could mean a possible loss of jobs and replacement of human reporters, as well as the spread of misinformation and lack of direct contact with the audience.
LatAm Journalism Review (LJR) looks behind-the-scenes at each initiative and talks to their creators.
Grupo Fórmula introduced Nat as the first AI-created anchor in the region. Her creator, Oswaldo Aguilar, explained told LJR that Nat is the product of a combination of several technologies and that it took four different AI softwares to bring her to life.
"For the animation we use a solution specially designed for us, we use another independent AI that produces her voice and another one to create the avatar. We use Chat GPT to create the stories she presents, but under two important considerations: First, we do not allow her to search for news stories outside of Radio Fórmula's corporate portal. Second, the stories are 'curated' by human intelligence that verifies that the information is truthful," Aguilar said.
Nat is slim, has dark hair and Latin features. She has interacted with human anchors in the media group where she has made it clear that she is not looking to replace them. "I’m not here to replace the work of any anchor. On the contrary, I am here to contribute original content and innovate in the technological area," Nat said in her presentation to the audience.
Nat's arrival has been experienced as a media event and comments on social networks have been varied. "Many news outlets have sought us out to talk about her. As for people's comments, they vary according to the social media outlet. On LinkedIn, they are all positive and on Facebook, there are many negative ones that reflect fears and prejudices about AI. Within the media group, everyone has reacted very well and wants to interact with it," Aguilar said.
Grupo Fórmula also decided to create Max, Nat's male counterpart, who focuses more on sports and financial stories. A newscast hosted by the two is expected to be launched soon. Currently, both Nat and Max can be seen on the media network's multi platform applications.
In Venezuela, the use of avatars generated by artificial intelligence to disseminate news favoring Nicolas Maduro's regime has been heard about since February.
It started with some videos on YouTube as part of a new newscast, in English with Spanish subtitles, called "House Of News en Español." There, two virtual journalists, named Noah and Daren, created with Synthesia.io software, fed the narrative that there is a new economic boom in Venezuela that has brought welfare and stability to its citizens.
The Chavismo regime denied having anything to do with these videos and attacked the correspondent of the newspaper El País in Venezuela, Florantonia Singer, for the publication of an article titled: “They are not journalists, they are avatars: Chavismo promotes propaganda made with artificial intelligence.”
"The same day that article appeared in El País, the Minister of Communications Freddy Ñañez wrote to me by Whatsapp sending me a thread of tweets making fun of the article. It is a bit heavy to have a minister writing to your phone forwarding his posts on Twitter where he questions you," Singer said.
"The issue was also discussed in an address by Nicolás Maduro, where he said that in El País we were lying. And then, in the Hojilla of Mario Silva [an opinion program at the State channel], a whole show was made to make fun of the story and talk about me and my family. The fact that he mentioned my family on national television was a breaking point. It shows how precarious the legal framework in our country has become. Undoubtedly, one is becoming more and more vulnerable," the journalist said.
A few weeks later, Nicolás Maduro introduced Sira [a name that resembles Siri, Apple's famous assistant], a brunette virtual anchor with long curly hair, created under the parameters of artificial intelligence, on his television program.
"Humanity has entered a new era and we have to be at the forefront in every way," Maduro said.
The Venezuelan organization ProBox, which seeks to combat digital disinformation, has followed up on this issue and says that no technology "is bad" in and of itself, but it depends on the use it is put to. "While in most of the world (at least in democratic countries), artificial intelligence has been associated with technological developments in favor of improving the lives of human beings; it has been used to improve the functioning of medical equipment, predict unfavorable environmental scenarios, help alleviate workloads in an automated way and even to optimize forms of learning in educational spaces, etc. In Venezuela, like many other technologies mainly designed for human development, contrary to this, it has been used by the government to promote disinformation and its propaganda in different spaces," Marivi Marín Vásquez, political scientist and executive director of ProBox, told LJR.
"I believe that the main intention of Chavismo with the use of artificial intelligence is, on the one hand, to use sophisticated strategies to amplify their narratives in different spaces and, in general, to continue using all available resources at their disposal to manipulate the information they cannot directly control," added the political scientist.
The School of Arts and Human Sciences of the Universidad Nacional Mayor de San Marcos in Peru had been exploring artificial intelligence for some time. With the advent of software such as Chat GPT and Dall-e, they decided to take a leap forward and create Illary, an anchor with Indigenous features who speaks Quechua [a language native to the Peruvian Andes].
Back in February, the university had already brought the famous Peruvian writer and anthropologist José María Arguedas to life thanks to AI in a commercial. This time, with Illary, they seek to innovate and return to the screens the first newscast created in Quechua that has been running since 2019.
The creators of Illary have told the media that Dall-e and DI-D were the programs used in this project. The phonetic templates were provided by Professor Óscar Huamán Águila, Quechua chair at San Marcos, who was previously in charge of narrating the newscast.
With the advent of these technologies, the debate about the possible loss of journalists' jobs has been opened. There are those who never thought that fiction would surpass reality and others who are more concerned about disinformation and the loss of contact with the audience.
"Artificial intelligence is indeed going to take away some jobs, but what worries me is that this is accelerating too fast and there is no margin to educate audiences," Singer said.
"We've already come from a pandemic of misinformation and a lot of journalists have had to go after fake news and disprove them. And now, with artificial intelligence, I think we need to educate audiences more than ever about the role of journalists and journalism in society. It is time to roll up our sleeves to go after this, without falling too far behind."