Debunked: Viral Image of Ancient Greek Mosaics Proven to be AI-Generated

A captivating image purporting to depict recently unearthed mosaics from ancient Greece has taken social media by storm, sparking excitement among history enthusiasts. The image, widely circulated on platforms like Facebook, showcased vibrantly colored mosaics featuring human figures and intricate patterns, supposedly dating back to the 2nd century BC. However, expert analysis and rigorous fact-checking have definitively exposed this image as an AI-generated fabrication.

Dr. Kutalmis Görkay, the archaeologist erroneously linked to the discovery in the viral posts, categorically denied any involvement with the depicted mosaics, labeling the circulating information as "fake news." The director of the Zeugma Mosaic Museum, a renowned repository of ancient mosaics, further confirmed that the artwork in question was not part of their collection. Görkay, upon closer inspection of the image, identified telltale signs of AI generation, citing inconsistencies in the mosaics’ style and scale.

Leading archaeology experts corroborated Görkay’s assessment, highlighting numerous historical inaccuracies within the image. Ruth Westgate, a senior lecturer in Ancient History and Archaeology at Cardiff University, pointed out the anomalous appearance of the mosaics. They lacked the characteristic tesserae (mosaic cubes) used in authentic Roman mosaics, and their placement, isolated shapes in the middle of the room, deviated from the standard practice of covering the entire floor. Westgate further observed the oddly proportioned and flat figures, a stark contrast to the naturalistic and painterly style of genuine Zeugma mosaics. The unusual figure with a giant eye, standing on human legs, was particularly perplexing and unlike any known ancient mosaic.

Professor Emerita Katherine Dunbabin of McMaster University echoed Westgate’s skepticism, labeling the image an “absurd fabrication.” She emphasized the overtly modern appearance of the mosaic imagery, lacking any semblance of Greco-Roman artistic conventions. Dunbabin even cast doubt on the image’s background, suggesting that an old photograph, predating the construction of the Birecik Dam, was likely used to create a misleading setting on the Euphrates River. The area depicted in the image has been submerged since the dam’s completion in 2000.

Further solidifying the debunking, AI detection tools were employed to analyze the image. Professor V.S. Subrahmanian of Northwestern University subjected the image to five deepfake detection algorithms. While two models indicated a probability above 0.5 for AI generation, they couldn’t provide absolute certainty. However, a human analyst noted several red flags suggesting inauthenticity, including the deformed and disproportionately sized individual depicted, and the lack of an expected shadow given the lighting conditions. Additional inconsistencies within the mosaic itself, such as a deformed hand, an incomplete head, and a partially missing hand, further contributed to suspicions of manipulation.

Olga Papadopoulou, a research assistant at the Centre of Research & Technology Hellas, also analyzed the image using a suite of synthetic image detection tools. One model, SPAI, registered a staggering 99.99% probability of synthetic origin, while two others indicated synthetic probabilities of 68.28% and 68.87%, respectively. Despite these compelling findings, the original source of the fabricated image remains elusive.

This incident serves as a potent reminder of the increasing sophistication of AI-generated imagery and its potential to spread misinformation. The rapid dissemination of this fabricated image underscores the importance of critical thinking, fact-checking, and consulting expert opinions when encountering potentially dubious content online. Dr. Görkay’s intention to file a formal complaint with the Turkish Ministry of Culture and Tourism highlights the seriousness of this issue and the need to combat the spread of false information masquerading as archaeological discovery. While AI can offer incredible creative possibilities, its capacity for misuse requires vigilance and a discerning approach to online content consumption.

Share.
Exit mobile version