AI-Generated Deepfakes Raise Election Disinformation Fears, But Impact Remains Theoretical
-
AI tools can quickly generate realistic but false images and audio to undermine elections through disinformation. Examples include fake photos of dumped ballots or sick candidates.
-
There are fears these AI-generated "deepfakes" could flood social media and sow doubts around the 50+ elections happening in 2024. However, impact so far seems mostly theoretical.
-
Deepfakes have targeted individual candidates, like a fake pornographic video of a Northern Irish politician. Some national leaders have been shown realistic deepfakes of themselves to highlight the threat.
-
Major tech companies like Meta and Microsoft have pledged to combat harmful uses of AI in elections through efforts like labeling AI-generated content. But most governments lack technical expertise to address this.
-
While AI enables faster, cheaper disinformation, clear evidence it has directly changed voter behavior or election outcomes is frustratingly elusive so far. Impact remains more potential than actual.