Two days before Slovakia’s September 30, 2023 parliamentary election, a fake audio clip spread on social media purporting to capture Michal Simecka, leader of the pro-European Progressive Slovakia party, and journalist Monika Todova of the newspaper Dennik N discussing how to rig the vote by buying votes and manipulating the count. Both quickly said the recording was fabricated, but it circulated widely on Facebook and Instagram, amplified by timing: it landed during Slovakia’s pre-election “silence period,” when media and candidates are restricted from responding to such claims.
A peer-reviewed analysis published in the Harvard Kennedy School Misinformation Review in August 2024 by Lluis de Nadal and Peter Jancarik examined “the Slovak case,” which had been widely billed as possibly the first election swung by a deepfake. Their conclusion was more measured. They argued that attributing the result to the deepfake overstates its effect and ignores the deeper factors that made parts of the Slovak electorate receptive to pro-Russian, anti-Western messaging, including low institutional trust and a history of conspiracy-prone narratives. The fake was real and notable, but it was one input into an outcome with many causes.
The episode became a touchstone in debates about AI and elections precisely because of this tension: a genuine, well-timed synthetic media attack, and a temptation to credit it with more power than the evidence supports.
For a business reader, it shows both sides of the deepfake threat: synthetic audio is now cheap and convincing, but panic about its impact can itself distort decisions, so claims about effect deserve the same scrutiny as the fakes themselves.