[Fig. 1.  Side-by-side stills contrasting the deepfake Zelensky video on the left with a genuine video of President Zelensky on the right making remarks at a news conference days prior.]

As of the writing of this article the video has had no discernible direct impact on the battlefield or Ukraine’s war effort, likely due to its relatively poor quality.  But the confusion it sowed, even if temporary, provided immediate and worldwide effects in the information space and demanded priceless time and attention from President Zelenskyy and members of his administration to rebut.

The episode remains a clarion call to those who contemplate the future of media manipulation and digital deception.  The evolutionary march of digital deception leads straight to the battlefield, and few capabilities when at their highest potential are better primed to cause confusion and chaos in the battlefield’s information space than deepfake technology.

“Deepfake” is the term associated with ultra-realistic video and audio images created not by human actors but by artificial intelligence.  Originally associated with salacious pornography videos that depicted unwitting victims participating in sex acts, people have used the technology to create perceptually perfect fake videos of such figures as President Barack Obama, celebrities like Emma Watson and Nicolas Cage, or even Russian President Vladimir Putin as early as 2018.  The technology has manipulated images of weather patterns and even depicted the life cycle of a daisy without needing human input for guidance.

The Zelenskyy deepfake is also not the first time that a deepfake has made a mark during a time of crisis.  In 2019, a deepfake-caused crisis instigated an attempted coup in Gabon, which nearly caused a civil war.  Supporters of Gabonese President Ali Bongo Ondimba became convinced that, after the President had not been seen for several days, a video purporting to show President Bongo alive, astute, and on the job was not real but instead was a deepfake. In support of this assumption, citizens pointed to differences in the President’s demeanor, physical appearance, his apparent inability to use a hand, and even raised skepticism about the video’s lighting.  Local newspapers had also speculated about deepfake, and on January 7, 2019, military officers from the Gabonese armed forces attempted a coup d’état by forcibly seizing a broadcast station and sending messages in an effort to “restore democracy.”  

While the coup did not succeed and the video was most likely real, the impact of the episode is enough to give skeptics of deepfake manipulation further pause.  No actual manipulation was necessary.  Deepfake technology’s existence alone brought the country to the edge of non-international armed conflict.  

With media manipulation at such new heights, international actors must not neglect its technical and legal impact on the battlefield.  This Article therefore attempts to assess the current state of deepfake technology, look ahead to its potential future applications in armed conflict, process the ways in which current law contemplates such deception, and distill recommendations for improving governance where needed.