Generative AI Risks Further Atomizing Democratic Societies

In early October 2023, Michal Šimečka, a progressive contender for the presidency in Slovakia, was defeated by his pro-Kremlin rival following the circulation of a fabricated audio clip online just days before the election. The clip, featuring a manipulated voice resembling his own, implied a willingness to engage in vote-buying tactics for victory. Shortly thereafter, in Argentina, the libertarian candidate Javier Milei secured the presidency amidst a campaign marked by the utilization of generative artificial intelligence (AI) to amplify his image and tarnish his opponents’. Milei depicted himself as a patriotic lion protector, a stark contrast to his public persona as an outspoken firebrand. Meanwhile, supporters of his opponent, Sergio Massa, employed AI to depict Massa as various heroic figures from Hollywood films while portraying Milei as despicable villains. Massa’s campaign even disseminated a fabricated video on Instagram, later removed, portraying Milei advocating for a legal market in human organs.

The era of deepfake campaigns has dawned, raising concerns about the erosion of social cohesion in the Western world. As generative AI becomes more prevalent and sophisticated, the proliferation of misinformation and deepfakes may lead to a troubling consequence: voter apathy. Faced with the daunting task of distinguishing truth from fiction in a digitally manipulated landscape, an increasing number of voters may opt out of political engagement altogether. Instead, they may seek refuge in the instant and endless escapism offered by generative AI programs, regardless of their authenticity.

Undoubtedly, anti-democratic forces are poised to exploit this potential outcome. Steve Bannon, former head of Breitbart News and a key figure in Donald Trump’s ascent to power, articulated a strategy in 2018 that is now emblematic of authoritarian populist movements worldwide: to manipulate the information environment to influence elections. As we approach 2024, a new wave of synthetic content is poised to inundate the internet. Nina Schick, a technology expert and author of “Deepfakes: The Coming Infocalypse,” predicts that the generative AI arms race among tech companies, spurred by innovations like ChatGPT, could lead to as much as 90 percent of online content being artificially generated by 2025. While the exact trajectory remains uncertain, the electoral cycles of 2024, when a quarter of the global population is eligible to vote, will likely provide a preview of the AI-infused political landscape looming on the horizon.