Multi-award-winning art director and visual journalist Adonis Durado (inset) and the AI recreation of a fire in Tondo, Manila.

An Expat Media exclusive with the multi-award-winning visual journalist on the power and perils of AI

When Adonis Durado, a multi-award-winning art director, visual journalist and former Dubai expat, decided to experiment with AI technology, he didn’t anticipate the chilling results. Driven by curiosity, Durado used Generative AI to recreate scenes from the massive fire that recently ravaged Isla Puting Bato, a slum in Tondo, Manila. The outcome? Hyper-realistic images so convincing they could easily pass as authentic photographs. The results left him shocked—and deeply concerned about the implications of this technology. “I watched the news about the fire. It was heartbreaking—residents scrambling to escape, some swimming for their lives. Seeing photos of the tragedy circulating online, I decided to see how Generative AI would interpret the event,” Durado shared. Using MidJourney, a leading AI tool, he generated scenes that mirrored the devastation with eerie precision. The images produced by the AI were startlingly realistic. The chaotic shanties in flames, the smoke-filled air, and the desperation etched on the faces of the people all seemed to capture the grim reality of the tragedy. But none of it was real. “The results stopped me in my tracks. The images looked so realistic that, without a second glance, they could easily be mistaken for actual photographs,” Durado said. “In the past, AI visuals had telltale flaws—distorted hands, unrealistic features—but now, the imperfections are almost gone.” Durado said today’s AI-generated visuals have reached a level where distinguishing fact from fiction has become a challenge. This experiment, however, was more than an artistic exercise. It raised troubling questions about how easily such technology could be misused. What if these images, stripped of context, were used to manipulate public perception or spread disinformation? Durado believes that the power of AI comes with an equally immense responsibility. He emphasized the danger of using AI to create deceptive visuals, particularly in the age of social media, where false narratives can go viral in minutes. “Social media is a breeding ground for emotional images. Many people don’t verify what they see; they just share. AI has the potential to blur the lines between reality and fabrication,” he said. “As creators, we must be transparent. Misusing AI erodes trust, especially in fields like journalism, where integrity is everything.” Durado’s concerns reflect the growing debate on AI’s ethical boundaries. While its creative potential is undeniable, the risk of it being used to peddle fake news or propaganda is equally significant. He urged the public to remain vigilant and to question the authenticity of what they consume online. Durado’s insights carry weight, bolstered by his remarkable career in the field of visual storytelling. An award-winning art director and visual journalist, he has spent most of his career working in Asia and the Middle East before moving to the United States. Currently a faculty member at Ohio University as a Knight fellow, Durado has an impressive résumé. He served as the design editor for Al Nisr Publishing in Dubai and as the design director of Muscat Media Group, where he oversaw two of Oman’s major daily newspapers. With over 100 international accolades to his name, including honors from the Society for News Design, the Society of Publication Designers, and the European Newspaper Awards, Durado has consistently been at the forefront of creative innovation in visual journalism. His freelance consultancy work has taken him across Africa, Asia, and Europe, making him a respected authority in his field.

Spotting AI-generated images: Durado’s expert advice

In an exclusive interview with Expat Media, Durado shared his tips for identifying AI-generated images. As someone who has extensively studied AI tools like MidJourney, DALL-E, and Adobe Firefly, he identified key signs that can help distinguish real from fake:

Human anatomy and faces:

AI often struggles with the details of human anatomy. Look for hands with extra fingers, misaligned pupils, open mouths with too many teeth, or overly smooth, waxy skin. These flaws are more pronounced in group photos or crowd scenes, where background faces may appear warped or incomplete.

Unnatural backgrounds and objects:

Common giveaways include floating objects, warped textures, and inconsistent lighting. Accessories like earrings may not align properly, and clothing can appear awkwardly rendered. The edges of objects or people are another weak point, often appearing blurry, jagged, or poorly defined.

Text elements:

AI-generated text, such as signs or book covers, often contains gibberish or distorted characters. Although newer tools like Ideogram are improving, text remains one of the easiest indicators of AI involvement.

Perspective and scale:

Pay attention to depth and spatial relationships. AI often struggles with proportions, resulting in misplaced or oddly scaled objects. For example, a giant spoon inexplicably sitting in a coffee cup or a bicycle missing key parts. Lighting inconsistencies are a common issue: shadows and highlights may not align with the apparent light source, and reflections in mirrors, windows, or sunglasses often appear distorted or nonsensical. “AI is evolving so quickly that even experts like me struggle to differentiate its output from reality,” Durado admitted. “Ironically, we may soon need AI tools to detect AI-generated content.” PIA/Expat Media

---

For all the latest news from the UAE and the world, follow us on Facebook, Twitter and Instagram and subscribe to our YouTube page