In the world where social media companies are working to crack down deepfakes and misinformation in the US Presidential Elections 2020, a politician in India named Manoj Tiwari used Artificial Intelligence techniques in videos that look like he said things that he didn’t say as reported by Vice.
In one version of the campaign video Manoj Tiwari speaks in English and in another fabricated video he speaks in the Hindi dialect of Haryanvi.
Political communications firm The Ideaz Factory said to the Vice that it was working with Bharatiya Janata Party to create “positive campaigns” using the same technology used in the deepfake videos.
“We used a ‘lip-sync’ deepfake algorithm and trained it with speeches of Manoj Tiwari to translate audio sounds into basic mouth shapes,” Sagar Vishnoi of The Ideaz Factory said, adding that it allowed the candidate to target voters he might not have otherwise been able to reach as directly because of different languages in different states.
According to Vice, the fabricated version of his video reached about 15 million people in India.
Deepfake videos are used to create nonconsensual pornography, but the now-infamous 2018 deepfake video of President Barack Obama raised major concerns about how false or misleading videos could be used in politics.
In October last year, California state passed a bill which made it illegal to share deepfakes of politicians and later in January, the US House Ethics Committee informed members that posting deepfakes on social media could be considered a violation of House rules.