Deepfake technology, leveraging artificial intelligence, enables the creation of fabricated videos depicting events that never occurred. This technology is well-suited for spreading disinformation, misinformation, and "fake news" on social media platforms, making it a potential tool in cyberwarfare.
Manipulated by AI, deepfake videos involve blending real and fake content, rendering them more realistic than videos generated solely by AI. For instance, deepfake technology can swap faces in a real video or alter lip movements to convey a different message. The rapid and accessible nature of deepfake creation has raised concerns about the ease with which fake videos can be disseminated.
One alarming example was a deepfake of President Zelensky falsely declaring the war's end, highlighting the potential for deepfakes, combined with compromised media services, to spread counterfactual messages from seemingly trustworthy sources. Similar deepfake videos of Russian President Vladimir Putin surrendering also emerged during the war.
Our research established a timeline of notable deepfakes early in the invasion and analyzed their discussions on Twitter. While some deepfakes were humorous or educational, they contributed to healthy skepticism. However, we observed instances of misinformation, where real videos were wrongly accused of being deepfakes, leading to a loss of trust in legitimate media.
The prevalence of deepfake videos online necessitates improved media literacy to balance healthy skepticism. Individuals should approach highly inflammatory media with caution, await verification from multiple trustworthy sources, and refrain from baseless accusations of videos being deepfakes.
As global conflicts unfold, the issue of deepfake misinformation will remain a significant concern, emphasizing the importance of media literacy in navigating an era where authenticity is increasingly challenged.
