QCOSTARICA — With the right Artificial Intelligence (AI) tools and the necessary technical knowledge, anyone could muddy the 2026 presidential campaign with “deepfakes.”
These are images, videos, audios and other multimedia content generated through these new technological tools, which have enormous potential for productivity and education, but also to generate false content and scams.
The warning is made by political strategists, deputies, possible presidential candidates, AI experts and university professors, who conclude that a significant percentage of Costa Ricans would be in diapers when it comes to detecting false content or deepfakes, while there is a huge legal vacuum to sanction behaviors that affect the reasoning of the vote due to these electronic tools.
– Advertisement –
“On this issue, there will be people who question the content of social networks from the first entry, but also citizens who do not. Society is not prepared to identify what is false and what is real. “Education and awareness campaigns are needed about the danger posed by these tools,” said Ariel Ramos, a professor of systems engineering at Universidad Fidélitas.
In recent days, this issue gained notoriety when presidential candidate Luis Amador reported that his voice was cloned in an audio that was leaked to the press and in which he and other people discussed a plan to declare President Rodrigo Chaves “mentally unstable.”
El candidato #LuisAmador niega la veracidad del siguiente audio que fue filtrado a la prensa en días pasados y en el cual, él y otras personas discuten un plan para declarar “mentalmente inestable” al presidente @rodrigochavesr #SoundCloud https://t.co/EPevmKIaxA
— Esteban Arrieta (@estarrar) September 17, 2024
With the emergence of ChatGPT in November 2022, AI has seen rapid growth and has entered basically all fields of productivity.
As part of their analysis and discussion, companies such as Meta and OpenAI, as well as non-governmental organizations and various users, have generated deepfake content with ridiculous connotations, but that seem real, such as Pope Francis stealing from a supermarket, to warn of the dangers of technology.
Usuarios en redes sociales han creado contenido de #InteligenciaArtificial en donde personajes como @realDonaldTrump y @KamalaHarris son puestos en situaciones ridículas, pero que parecen reales pic.twitter.com/qajuaIqonA
— Esteban Arrieta (@estarrar) September 17, 2024
– Advertisement –
Similarly, the RepresentUs group in the United States created videos of Kim Jong-un, leader of North Korea, and Vladimir Putin, of Russia, to expose the dangers of AI and how fragile a democracy can become.
More recently, singer Taylor Swift had to deny an AI-generated video circulating on Instagram in which she allegedly endorsed Donald Trump for president.
“To counteract this situation, a lot of education and awareness is required from all actors, who would have to come out and promptly refute the false messages that are generated,” said Gabriel Silva, an expert in Artificial Intelligence and Cybersecurity and member of the National Chamber of Information and Communication Technologies (Camtic).
Under current electoral legislation, not even trolls can be sanctioned, according to the Supreme Electoral Tribunal (TSE), since the law does not establish trolling or the creation of false accounts as an illicit conduct.
– Advertisement –
In this way, a person or an organization could attack with complete impunity and would only commit a crime if their conduct falls within the reasoning for identity theft established by the Penal Code.
“The AI law that we are drafting does not include anything on this topic, because we think that it is a very broad topic that must be addressed with greater care. It is undoubtedly a worrying issue that requires specific legislation,” said Vanessa Castro, a PUSC deputy and proponent of the project.
Concerns
Presidential candidates, analysts, technology experts and other actors refer to the challenges that the country will face due to the use of Artificial Intelligence in the political campaign.
For those consulted, it is practically a fact that the “deepfake” or identity theft through videos and audios will be added to the fake news that circulate on social networks during the electoral season.
“In recent days I was the victim of an irresponsible and orchestrated campaign to continue dirtying my name. On the subject of the leaked audios, I want to make it clear that I do not recognize the legitimacy of the audios where my voice is supposedly heard.
“What is evident is that their disclosure in a national media outlet related to the government is due to an irresponsible campaign to continue dirtying my name. I will not lend myself to this type of dirty politicking that contributes nothing to the Costa Rican citizenry,” were the words of Luis Amador, former minister of the MOPT and currently independent presidential candidate.
For Ivan Barrantes, political strategist, “If we do not do something with the education system, which was once of quality and generated excellent human resources, we will have thousands of people “cannon fodder” to be deceived by what is known as “deepfake”. Education is essential to create analytical and questioning citizens.”
“However, we must also be clear that Artificial Intelligence is advancing very quickly and it is possible that in the future, even the most prepared will be scored on. Today, the field is completely free and it is possible that in 2026 there will be something of this; otherwise, in 2030 for sure,” concluded Barrantes.
Some tools
Just by doing a Google search, you can find countless options for tools to generate content with Artificial Intelligence.
As technology advances, new options will become available.
These are some of the most popular for generating multimedia content. With Artificial Intelligence, you can make “deepfakes,” which use deep learning, an area of artificial intelligence, to obtain fake videos:
- Deepfacelab
- FaceSwap
- Vidnoz Face Swap
- Miocreate
- Voice Engine, is is a program that could be used to clone voices and make them look like originals
These are generative Artificial Intelligence programs and services that develop artificial intelligence:
- Midjourney
- DALL-E by OpenAI
- Diffusion by Stability
- Da Vinci
- Stable Diffusion
- Artiphoria
- Adobe Firefly
– Advertisement –
Source link
Rico