BUDAPEST: A video of Hungarian soldiers departing for Ukraine and returning in caskets is among numerous AI-generated posts appearing on Prime Minister Viktor Orban’s social media accounts ahead of next year’s elections.
Orban, a nationalist leader in power since 2010, faces an unprecedented challenge from opposition leader Peter Magyar in the upcoming contest.
Politicians and supporters across the political spectrum have adopted the technology, with Orban targeting the opposition following US President Donald Trump’s example.
Trump has repeatedly posted fabricated videos that stir online outrage.
Experts warn that emotionally charged AI content could sway public opinion before the poll.
Orban’s supporters have invested heavily in AI technology ahead of the April election.
The National Resistance Movement, which backs Orban’s Fidesz party, has targeted Magyar with unlabelled AI-generated videos on social media.
Since June 15, the group spent over 1.5 million euros on Facebook and YouTube promotions, more than any EU political organisation according to Meta and Google data.
Both US tech giants recently ended political advertising in the EU due to new rules requiring clear labelling and sponsor identification.
Some NEM videos feature non-existent people expressing outrage at Magyar’s alleged policies, while others are deepfake videos of the opposition leader.
Other pro-Orban AI clips support his policy positions, including opposition to Ukraine’s EU membership bid.
The AI-generated soldiers in caskets video was advertised on Facebook and broadcast on state TV during mail-in voting on Kyiv’s EU accession.
The government maintains that using such content is acceptable when properly labelled as artificially generated.
AFP found multiple instances of Orban and allies posting realistic-looking videos and pictures without clear AI labelling.
Magyar has also used unlabelled AI on his social media channels, including a fake photo of himself in combat uniform.
Magyar told AFP that clearly visible AI use is acceptable, while anything else could constitute election fraud.
He described Orban’s AI videos as pathetic.
The government’s AI commissioner Laszlo Palkovics said avoiding AI content to influence voters would be advisable.
Palkovics told AFP that everyone must decide their AI use while insisting government fully supports the EU AI Act.
Hungary plans to develop its own AI ethics code as Orban strives to make the country a key regional AI player.
Economy minister Marton Nagy admitted at a recent AI conference that local companies lag behind most EU peers in AI adoption.
Experts express concern about AI-generated content proliferation.
Communication researcher Petra Aczel told AFP that deepfake videos can have heightened impact with strong sensory influences expected in global political campaigns.
She added that even labelled AI content with visible glitches can be believed due to emotional reactions.
Hungarian-American AI consultant George Tilesch said the EU AI Act will not restrain political actors.
The Silicon Valley-based expert noted social media algorithms feed users similar content for maximum engagement.
He warned that combined with generative AI breakthroughs, this creates a potentially highly toxic concoction. – AFP
