00:00Every political ad behind me is completely fake.
00:03Some AI-generated content is obvious.
00:05It's intended as satire to entertain or trigger you.
00:08The three ads behind me have amassed more than 5 million views alone.
00:12But some AI ads are so sophisticated that they're almost impossible to detect.
00:17It's a danger to democracy, and we're already seeing the consequences.
00:21This video of anti-Trump Republican Thomas Massey was shared widely
00:25during Kentucky's 2026 Republican primary.
00:28His opponent created the ad using generative AI,
00:31but many older voters believed it.
00:34In fact, Massey told me he partly blames the ad for losing to his Trump-backed opponent.
00:39What's even worse is that tech companies sometimes can't detect
00:42the fake content that their own technology has created.
00:45In 2024, an audio deepfake of then-President Biden urging people not to vote
00:51was rated by Eleven Labs' detection tool as having just a 2% likelihood of being synthetic.
00:57But another voice fraud detection company determined the audio was in fact faked,
01:01and it had been created using, get this, Eleven Labs' own technology.
01:06Although media reports suggest that Republicans have so far leaned into AI-enhanced ads more
01:12than Democrats, members of both parties have deployed them.
01:14The more political communication becomes saturated with synthetic images and audio,
01:19the more likely the public will dismiss all political messaging as lies.
01:24That kind of distrust could lead to a cascade of problems,
01:27like increasing polarization and declining rates of participation in our democracy.
01:32But there are things we can do.
01:34I wrote a column exploring all of this and how AI is taking over political communication.
01:39You can read it on Bloomberg.com.
Comments