Artificial intelligence (AI) is the fourth of the great advances of the last half century to revolutionize communications, as important as the invention of the printing press 600 years ago: personal computers, the Internet and smartphones came before it. Today’s teenagers have no idea what the world of typewriters, telephone operators and public libraries was like (not to mention English stone calligraphy!).

AI, like previous inventions, is accompanied by the promise of a breakthrough in productivity that will bring prosperity and better living standards. The first generation of ChatGPT and its ilk is like a very smart, diligent, but novice research assistant: good at finding sources and data instantly, but unable to distinguish right from wrong or make independent judgment. ChatGPT would make a bad columnist.

Like any technological advance, AI carries dangers. The main ones for the current election process are:

– Qualitative increase in the ability to manipulate information and spread fake news. What used to be done with pamphlets, is now done by trolls who flood social networks with misinformation, especially Twitter. But with artificial intelligence, the manipulation is enhanced. Instead of trolls sending generic messages through social networks, the gigantic capacity of artificial intelligence will allow these programs to send personalized messages to each citizen, according to their profile, with the aim of influencing their electoral behavior. For example, a candidate can tell teachers that their salaries will increase, and young people that they will be more flexible when enrolling in colleges.

– AI can collect personal data and political preferences of citizens and use them to influence voter decisions. For example, if someone in the family suffers from a catastrophic illness, the candidate will send a message: that he intends to increase state funding for such illnesses.

– Manipulation of election results. Previously, only electoral authorities could change results, as was suspected in Ecuador in the past decade and routinely in Venezuela. But now hackers and malicious actors could use the automation of certain election processes, such as vote counting, to change the results, manipulate the will of the voters, and change the course of the election.

In order to counter these actions, which will take place, it is necessary to establish clear standards for the use of artificial intelligence, guarantee the participation of information security experts in all stages of the process, and warn citizens of the risks they are exposed to. A more long-term action is the inclusion of news content analysis in the school curriculum. Finland is leading in this discipline.

Abuse of artificial intelligence can threaten the integrity of the democratic process, undermine citizens’ privacy and undermine trust in election results. Electoral bodies are challenged to stay one step ahead of those who want to resort to this new technology to falsify the evaluation of the electorate. (OR)