Social Media decided last elections. Will AI decide the next ones?
AI image of Donald Trump fighting Joe Biden. Created by Frédéric De Vries with Midjourney.
You may remember the headlines. Brexit happened because of Facebook. Twitter got Trump elected. Then there was the Cambridge Analytica scandal that allegedly massively influenced the American elections in 2010. Closer to home, the president of a major political party warned me 'we' would win or lose the elections on social. We, as in the advertising agency that was about to run their campaign on social. While most of these statements are at least exaggerated, the influence of social media on elections worldwide is undeniable. Now it seems AI will play a significant role in the coming elections as well.
Argentina: the first AI elections
10 days ago, far-right outsider Javier Milei won the elections in Argentina. These elections will probably go down in history as the first AI elections. The campaign teams of both candidates made extensive use of AI in different ways. Some quite innocent, some downright worrisome.
Both candidates used AI tools to generate their campaign posters and images for their social channels. While Sergio Massa opted for a style that was clearly inspired by the visual language of Soviet propaganda, Javier Milei's team took it one step further to use it against him and depict him as a communist dictator.
Sergio Massa's Campaign Posters - source Instagram
Javier Milei's team depicting Sergio Massa as a Communist Leader - source Instagram
There is hardly any harm in prompting AI tools to create your own images this way. Instead of Photoshop and Illustrator, the campaign team nowadays uses AI tools that are faster and easier to use.
You could argue that it's less self-evident to use it it to manipulate images of your opponent. While it is clearly a 'fake' image above, as in : it's not a manipulated photograph, you might think that it's an image from Massa's campaign. That is already a bit problematic.
Next, Massa used deepfake video's of himself in heroic situations: a soldier in WW1, or as a ghostbuster. While clearly fake in these instances, one might wonder whether deepfake videos have a place in a political campaign. Because, where will you draw the line between 'clearly exagerated' (as in : depicting yourself as a soldier in WW1) or 'maybe possible' (as in: depicting yourself as a soldier in the Falkland war).
However, Massa's campaign team took it a couple of steps further.
At one point, The Massa campaign produced one "deepfake" video in which Mr. Milei explains how a market for human organs would work, something he has said philosophically fits in with his libertarian views.
"Imagine having kids and thinking that each is a long-term investment. Not in the traditional sense, but thinking of the economic potential of their organs," says the manipulated image of Mr. Milei in the fabricated video, posted by the Massa campaign on its Instagram account for A.I. content, called "A.I. for the Homeland."
I don't think there can be much debate about that.
Implications for Future Elections
The use of AI in Argentine elections raises obvious questions for the future of political campaigning. Some immediately come to mind.
Will AI become a decisive factor in winning elections, surpassing the impact of social media?
How will AI transform traditional campaigning strategies? Will it lead to more innovative and engaging voter outreach, or will it amplify the risks of misinformation and manipulation?
What ethical considerations arise with the use of AI in elections? How can we ensure a balance between leveraging technology for political advantage and maintaining democratic integrity?
As AI technology continues to advance, its integration into election strategies is inevitable. This integration will require some careful thinking of legislators. Here are three pillars that I think are important.
Regulatory Frameworks: Developing comprehensive regulations to govern the use of AI in political campaigning. I believe we should seriously think about whether or not we want to allow deepfakes.
Content transparency: Indicating whether AI was used to create materials might be a start: full transparency via content credentials might be better.
Public Awareness: Educating voters on AI's role in elections, helping them discern between genuine and AI-generated content.
Conclusion
To this very day, policy makers all over the world are debating about the influence of social media on the elections. Often, they see social media as a threat to our democracy. And yes, it's an important debate. Fighting disinformation on social media is hard and the effects of echo-chambers towards extremism has been demonstrated over and over.
But I think the next elections might not just be influenced by social media. They could be swayed by AI.
So I suggest policy makers and political parties start thinking about this. Preferably before the next elections. And hopefully far beyond them.
#Elections #AI #DigitalCampaigning #EthicalAI #PoliticalStrategy