AI-generated images are becoming a ubiquitous part of the internet. In Luxembourg, too, large institutions, charitable organisations and political parties are beginning to use deceptively real-looking but computer-generated images. Where are the limits? And can "fake" images be more ethical than "real" ones?
This article is provided to you free of charge. If you want to support our team and promote quality journalism, subscribe now.
Can we still trust our eyes? If you pay close attention when scrolling, you will have noticed a new trend in the feed: more and more "AI images" are appearing, i.e. images that have been created by computer programmes but sometimes look deceptively real. Between automatic filters and selfies that have been through Photoshop, "photos" online are no longer a reflection of reality. For a long time, images were, if not definitive proof, then at least an indication that what they depict actually happened (roughly) like this. Not any more. "Photos" of impressive world records (such as the world's largest tiramisu), incredible archaeological finds (like bones of giants), fantastic products at top prices, or unbelievable, colourful animals are mixed in between real snapshots and news.
Here a finger too much, there a jacket that fuses with the shirt on closer inspection, or text in the background that looks plausible but is gibberish. The technology that (simplified) combines elements from a large collection of images to create a plausible new image is far from perfect. But it is improving almost daily. It's certainly enough for scammers to unscrupulously collect a few "likes", deceive naive customers or spread false information. It's not really surprising that it has become a favourite tool for spreading misinformation, disinformation and conspiracy theories.
Continue reading for free
Get access to this article by subscribing to our newsletter that is sent twice a week. You also have to have a Journal account.
Already have an account?
Log in
