The Rise of AI: Political Ramifications and Future Challenges

Politics is one of the largest and most influential industries worldwide currently at risk due to the use of Artificial Intelligence.

The use of artificially generated content in campaigns can generate miscommunication and misinformation increasing the risk of problems for decisions, candidates and government policy proposals nationally and internationally.

In April 2023, President Joe Biden of the United States launched his re-election campaign by video, which, on the same day the Republican National Committee (RNC) responded with a 30-second advertisement including speculation that 4 years of President Biden’s term could include more crime, open borders, war with China and economic collapse, which generated discord and various comments on social networks including twitter and facebook, in the face of the first political attack after the campaign announcement.

The reality is that this was the first national public statement and political campaign generated entirely by images, voices and editing by Artificial Intelligence (AI) which was admitted by the RCN after being published. As a consequence, it generated a global threat and a warning about how artificial intelligence is influencing miscommunication and political manipulation worldwide.

Advances in digital technology promote new tools through virtual political messaging, impacting how political members, voters and reporters create a new perspective on campaigns and decisions.

In addition, artificial intelligence has the ability to find the right audience depending on the instruction of the content, which is fundamental in political campaigns, as candidates will use it to their advantage to increase votes and manipulate the final decision, using micro data and brokers that can encrypt the information that people read and see including their political preferences, in order to analyze the material in real time.

According to “Stanford University Human-Centered Artificial Intelligence”, a research conducted by professor of sociology and psychology, Robb Willer, in the Stanford School of Humanities and Sciences and director of the Polarization and Social Change Lab, they made use of GPT-3, the same language as GPT chat, asking it to speak persuasive messages on different persuasive topics.

After that, they had different people read these texts randomly, sometimes they were written by humans, and sometimes they were written by AI. In all cases, the participants were asked to state their problems before and after reading the article so that the research team had the opportunity to see how persuasive the messages had been on the readers and to identify which were most persuasive and why.

This led to the conclusion that the three comparisons said that AI generated messages that were consistently persuasive in the eyes of humans, and that humans were more supportive of the ideas that were produced by AI.

The team took the study as an act of caution about the possible next steps and influence that artificial intelligence will have, saying that chatbots have serious democratic and national security implications.

Concerns were also raised about how this can be a very dangerous weapon used in the political context, as long language models, such as GPT-3 and GPT-4, can be manipulated to use false or inaccurate information for political purposes.

AI has already concluded a level of sophistication that rises to a state of persuasion in law that requires considerable attention. It has the potential to change a political discourse in real time on television, which requires that the world’s major technology companies start working on moderators that can decrease the risk of these incidents, in order to prevent wars between countries and national discord between cities.

And even for a more distant future, it is possible that political positions such as mayor, councilor, senator, deputy, ministers, may possibly be replaced and complemented by AI, because it offers an immediate response and can make a study with statistics in real time to contribute to national decisions depending on the situation of local cities more effectively.

Author: Mariana Lamprea Andrade