Images of Donald Trump’s arrest circulating on social media are fake, however, they attest to the considerable progress being made in AI-generated images. Researchers say that looking outside the image’s focal point, examining the body proportions, and deciphering odd writing on walls and clothing are just a few ways of identifying fake images. The old tell of over-the-top facial expressions still works well as AI-generated images tend to render such expressions in an exaggerated way, with skin creases being very pronounced. While famous people are rendered very well, people of less stature are often depicted as “wonky”. Twitter’s policy on AI-generated images is to not share synthetic, manipulated, or out-of-context media that may deceive or confuse people and lead to harm. However, the rise of AI art is raising questions about social media platforms’ approach to synthetic content and the need to guide users through this complex and often unsettling world of generative AI.
Identifying deepfake AI images of Trump’s arrest is possible through visual enhancements and authenticity verification tools.
