The announcement by Joe Biden that he will not seek reelection has sparked widespread reactions online, particularly following his endorsement of Vice President Kamala Harris as the Democratic nominee.
Supporters of Harris took to social media to highlight humorous moments and memorable speeches from her time in office, including her quirky phrase, “Do you think you just fell out of a coconut tree?”
In contrast, supporters of Republican nominee Donald Trump have circulated manipulated content, including a deepfake video that depicts a speech never actually delivered by Harris.
This deepfake, which has gained significant traction on social media, shows Harris addressing a crowd, but it features altered video and audio generated by AI technology.
A report released revealed that the deepfake has garnered millions of views. Following its exposure, social media platforms took swift action to remove the misleading content.
In a statement, a spokesperson emphasized the platform’s commitment to eliminating harmful AI-generated content and collaborating with fact-checkers to verify the accuracy of posted materials.
Resurgence of Harris Deepfake Following Presidential Candidate Announcement
This is not the first instance of the Kamala Harris deepfake circulating online. Previous evaluations have disproven the clip, which was first shared last year. The altered video consists of genuine footage of Harris speaking at a Howard University event.
In the viral deepfake, Harris bizarrely states, “Today is today and yesterday was today yesterday,” although she never made such a statement.
Please note that the full recording of the actual event does not contain the manipulated quote. Digital experts have highlighted inconsistencies in the deepfake, such as unnatural noise around Harris’s mouth, indicative of editing efforts to synchronize the visuals with the fake audio. Additionally, the audio lacks ambient background sounds typical of a live event.
More than a year after its initial debunking, this Harris deepfake resurfaced on social media, achieving over 3.4 million views and remaining on the platform despite its policy against misleading content. Users have made efforts to add clarifications indicating the video’s inaccuracy.
In contrast, another social media platform has dedicated resources to identify and remove misinformation. It reports a high success rate in removing violating content but stated that one viral deepfake of Harris acquired over 4.1 million views before being taken down. The platform is actively working to identify and eliminate other instances of the deepfake.
With the increasing availability of AI-generated audio and video tools, deepfakes are anticipated to pose a significant challenge in upcoming political campaigns, making their impact in 2024 a noteworthy concern.
